GENERAL STANDARD FOR CONTAMINANTS AND TOXINS IN FOOD AND FEED

CXS 193-1995
Adopted in 1995.
Revised in 1997, 2006, 2008 and 2009.
Amended in 2010, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2021, 2022, 2023 and 2024.
CXS 193-1995 2
Revision 2024

Following decisions taken at the Forty-seventh Session of the Codex Alimentarius Commission in November
2024, maximum levels of lead were added for dried aril, dried floral parts, dried fruits and berries, dried paprika
and sumac, dried Sichuan pepper and star anise, dried rhizomes and roots, dried seeds, dried celery seeds
and quinoa, and the maximum level of cadmium in quinoa. Annex IX Sampling plan for methylmercury
contamination in fish was added.
CXS 193-1995 3
1.1 SCOPE
This standard contains the main principles which are recommended by the Codex Alimentarius in
dealing with contaminants and toxins in food and feed and lists the maximum levels (MLs) and
associated sampling plans of contaminants and natural toxicants in food and feed which are
recommended by the Codex Alimentarius Commission (CAC) to be applied to commodities moving in
international trade.
This standard includes only MLs of contaminants and natural toxicants in feed in cases where the
contaminant in feed can be transferred to food of animal origin and can be relevant for public health.
1.2 DEFINITION OF TERMS
1.2.1 General
The definitions for the purpose of the Codex Alimentarius, as mentioned in the Procedural Manual of
the Codex Alimentarius Commission,1 are applicable to this standard and only the most important ones
are repeated here. Some new definitions are introduced, where this seems warranted to obtain optimal
clarity. When reference is made to foods, this also applies to animal feed in those cases where this is
appropriate.
1.2.2 Contaminant
Codex Alimentarius defines a contaminant as follows:
“Any substance not intentionally added to food or feed for food-producing animals, which is present in
such food or feed as a result of the production (including operations carried out in crop husbandry,
animal husbandry, and veterinary medicine), manufacture, processing, preparation, treatment, packing,
packaging, transport or holding of such food or feed, or as a result of environmental contamination. The
term does not include insect fragments, rodent hairs and other extraneous matter”.
This standard applies to any substance that meets the terms of the Codex definition for a contaminant,
including contaminants in feed for food-producing animals, except:
1) Contaminants having only food and feed quality significance (e.g. copper) but no public
health significance in the food(s) given that the standards elaborated within the Codex
Committee on Contaminants in Foods (CCCF) have the objective to protect public health.
2) Pesticide residues, as defined by the Codex definition, that are within the terms of reference
of the Codex Committee on Pesticide Residues (CCPR).
3) Residues of veterinary drugs, as defined by the Codex definition, and residues of feed
additives*, that are within the terms of reference of the Codex Committee on Residues of
Veterinary Drugs in Foods (CCRVDF).
4) Microbial toxins, such as botulinum toxin and staphylococcus enterotoxin, and
microorganisms that are within the terms of reference of the Codex Committee on Food
Hygiene (CCFH).
5) Residues of processing aids that are within the terms of reference of the Codex Committee
on Food Additives (CCFA)**.
(*) Feed additives as defined in the Code of practice on good animal feeding (CXC 54-2004):2 “Any
intentionally added ingredient not normally consumed as feed by itself, whether or not it has nutritional
value, which affects the characteristics of feed or animal products”.
Residues of feed additives include the parent compounds and/or their metabolites in any edible portion
of the animal product and include residues of associated impurities of the feed additive concerned.
(**) Processing aids are any substance or material, not including apparatus or utensils, and not consumed
as a food ingredient by itself, intentionally used in the processing of raw materials, foods or its
ingredients, to fulfil a certain technological purpose during treatment or processing and which may result
in the non-intentional but unavoidable presence of residues or derivatives in the final product.
1.2.3 Natural toxins included in this standard
The Codex's definition of a contaminant implicitly includes naturally occurring toxicants, including toxic
metabolites of certain microfungi that are not intentionally added to food and feed (mycotoxins).
Toxins that are produced by algae and that may be accumulated in edible aquatic organisms such as
shellfish (phycotoxins) are also included in this standard. Mycotoxins and phycotoxins are both
subclasses of contaminants.
CXS 193-1995 4
Endogenous natural toxicants, such as solanine in potatoes, that are implicit constituents of food and
feed resulting from a genus, species or strain ordinarily producing hazardous levels of a toxic
metabolite(s), i.e. phytotoxins, are not generally considered within the scope of this standard. They are,
however, within the terms of reference of CCCF and will be dealt with on a case-by-case basis.
1.2.4 Maximum level and related termsi
The Codex maximum level (ML) for a contaminant in a food or feed commodity is the maximum
concentration of that substance recommended by the CAC to be legally permitted in that commodity.
1.3 PRINCIPLES REGARDING CONTAMINANTS IN FOOD AND FEED
1.3.1 General
Contamination of food and feed may pose a risk to human (and/or animal health). Moreover, in some
cases, they may also have a negative impact on the quality of the food or feed. Food and feed can
become contaminated by various causes and processes.
Contaminant levels in food and feed shall be as low as reasonably achievable through best practices
such as good agricultural practice (GAP) and good manufacturing practice (GMP) following an
appropriate risk assessment. The following actions may serve to prevent or reduce contamination of
feed and food.ii
 Preventing food and feed contamination at the source, e.g. by reducing environmental
pollution.
 Applying appropriate technology control measure(s) in food and feed production,
manufacture, processing, preparation, treatment, packing, packaging, transport or holding.
 Applying measures aimed at decontamination of contaminated feed or food and measures
to prevent contaminated feed or food to be marketed for consumption.
To ensure that adequate action is taken to reduce contamination of food and feed, a code of practice
shall be elaborated comprising source-related measures and GMP as well as GAP in relation to the
specific contamination problem.
The degree of contamination of food and feed and the effect of actions to reduce contamination shall be
assessed by monitoring, survey programmes and more specialized research programmes, where
necessary.
When there are indications that health hazards may be involved with consumption of food that is
contaminated, it is necessary that a risk assessment should be undertaken. When health concerns can
be substantiated, a risk management measure must be applied, based on a thorough evaluation of the
situation and consideration of a range of risk management options. Depending on the assessment of
the problems and the possible solutions, it may be necessary to establish MLs or other measures to
control the contamination of food and feed. In special cases, specific advice on dietary recommendations
may also have to be considered to complement other regulatory measures, when the measures are not
sufficiently adequate to protect public health and safety.
National measures regarding food and feed contamination should avoid the creation of unnecessary
barriers to international trade in food and feed commodities. The purpose of this standard is to provide
guidance about possible approaches to eliminate or reduce the contamination problem and to promote
international harmonization through recommendations, which in turn may prevent trade barriers and
disputes.
For all contaminants, which may be present in more than one feed or food item, a broad approach shall
be applied, considering all relevant information that is available, for the assessing of risks and for
developing recommendations and control measures, including the setting of MLs.

i For the contaminants radionuclides, acrylonitrile and vinylchloride monomer a Codex guideline level (GL) has
been established.
A Codex guideline level (GL) is the maximum level of a substance in a food or feed commodity which is
recommended by the Codex Alimentarius Commission to be acceptable for commodities moving in international
trade. When the GL is exceeded, governments should decide whether and under what circumstances the food
should be distributed within their territory or jurisdiction.
Because the Commission has decided that the preferred format of a Codex standard in food or feed is a maximum
level, the present existing or proposed guideline levels shall be reviewed for their possible conversion to a maximum
level after a risk assessment performed by JECFA, if appropriate.
ii In addition, reference is made to the Code of practice concerning source directed measures to reduce contamination
of food with chemicals (CXC 49-2001) and the Code of practice on good animal feeding (CXC 54-2004).
CXS 193-1995 5
1.3.2 Principles for establishing maximum levels in food and feed
MLs shall only be set for food in which the contaminant may be found in amounts that are significant for
the total exposure of the consumer, taking into consideration the “Policy of the Committee on
Contaminants in Foods for exposure assessment of contaminants and toxins in foods or food groups”
(Section 4.5 of the Codex Procedural Manual1).
The MLs shall be set in such a way that the consumer is adequately protected. At the same time, the
other legitimate factors need to be considered. This will be performed in accordance with the Working
principles for risk analysis for food safety for application by governments (CXG 62-2007).3
The principles of GMP and GAP, as defined by Codex, shall be used. MLs shall be based on sound
scientific principles leading to levels, which are acceptable worldwide, so that there is no unjustified
barrier to international trade. MLs shall be clearly defined with respect to status and intended use.
1.3.3 Specific criteria
The following criteria should (not preventing the use of other relevant criteria) be considered when
developing MLs and/or other measures in connection with this standard (further details about these
criteria are given in Annex I).
Toxicological information
 identification of the toxic substance(s);
 metabolism by humans and animals, as appropriate;
 toxicokinetics and toxicodynamics, including information on possible carry-over of the toxic
substance from feed to edible animal tissue/products;
 information about acute and long-term toxicity and other relevant toxicity data; and
 integrated toxicological expert advice regarding the acceptability and safety of intake levels
of contaminants, including information on any population groups which are especially
vulnerable.
Analytical data
 validated qualitative and quantitative data on representative samples; and
 appropriate sampling procedures.
Intake data
 presence in food of dietary significance for the contaminant;
 presence in food that are widely consumed;
 presence in feed and feed components;
 food intake data for average and most exposed/high consumer groups;
 results from total diet studies;
 calculated contaminant intake data from food consumption models;
 data on intake by susceptible groups; and
 data on intake by food-producing animals.
Technological considerations
 Information about contamination processes, technological possibilities, production and
manufacturing practices and economic aspects related to contaminant level management
and control.
Risk assessment and risk management considerations (cf. Working principles for risk analysis for
food safety for application by governments)
 risk management options and considerations;
 consideration of possible MLs in food and feed based on the criteria mentioned above; and
 consideration of alternative solutions.
1.4 FORMAT OF THE GENERAL STANDARD FOR CONTAMINANTS IN FOOD AND FEED
A full description of the format is provided in Annex II.
CXS 193-1995 6
Annex I
CRITERIA FOR THE ESTABLISHMENT OF MAXIMUM LEVELS IN FOOD AND FEED
Introduction
In this annex, criteria are mentioned regarding information, which is considered necessary for evaluating
contaminant problems in food and feed and for the establishment of maximum levels (MLs). The criteria
mentioned here are elaborated in more detail than in Section 1.3.3 of the Preamble. Only those aspects that
need further clarification are detailed; however, criteria or aspects that are not specifically detailed here should
not be ruled out in the evaluation process.
Toxicological information
Integrated toxicological expert advice regarding a safe/tolerable intake level of a contaminant is essential
when decisions about MLs in foods are considered. A recommendation from the Joint FAO/WHO Expert
Committee on Food Additives (JECFA) regarding the maximum allowable or tolerable intake, based on a full
evaluation of an adequate toxicological database, should be the main basis for decisions by Codex members.
In urgent cases, it may be possible to rely on less developed evaluations from JECFA or on toxicological expert
advice from other international or national bodies.
When toxicological information is presented in relation to proposals for MLs for contaminants in food and feed,
information about the following aspects is desirable:
 identification of the toxic substance(s);
 metabolism in humans and animals, as appropriate;
 toxicokinetics and toxicodynamics including information on possible carry-over of the contaminant
from feed to edible animal tissue/products;
 information about acute and long-term toxicity in animals and humans, including epidemiological
data on humans and other relevant toxicity data;
 conclusions and advice of toxicological expert(s) (groups), with references, including information
on especially vulnerable population groups or animals.
Analytical data
Validated qualitative and quantitative analytical data on representative samples should be supplied.
Information on the analytical and sampling methods used and on the validation of the results is desirable. A
statement on the representativeness of the samples for the contamination of the product in general (e.g. on a
national basis) should be added. The portion of the commodity that was analysed and to which the contaminant
content is related should be clearly stated and preferably should be equivalent to the definition of the
commodity for this purpose or to existing related contaminant regulation.
Information on appropriate sampling procedures should be supplied. Special attention to this aspect is
necessary in the case of contaminants that may not be homogeneously distributed in the product
(e.g. mycotoxins in some commodities).
Intake data
It is desirable to have information about the contaminant concentrations in those foods or food groups that
(together) are responsible for at least half and preferably 80 percent or more of the total dietary intake of the
contaminant, both for consumers with average and high consumption patterns.
Information about the presence of the contaminant in foods that are widely consumed (staple foods) is
desirable in order to be able to make a satisfactory assessment of the contaminant intake and of risks
associated with food trade.
For the contaminants, which can be present in food of animal origin as a consequence of the carry-over from
feed, information about the presence of the contaminant in the feed and feed components should be given.
Furthermore, the intake of contaminants by the different food-producing animals and the resulting levels of the
contaminant in the food of animal origin should be estimated.
Food consumption data for average, most exposed (high consumers), and susceptible consumer
groups are desirable for evaluations of (potential) intake of contaminants. This problem, however, has to be
addressed differently on a national and on an international scale. It is therefore important to have information
about both average and high consumption patterns regarding a wide variety of foodstuffs so that for every
contaminant, the most exposed consumer groups may be identified for every contaminant. Detailed information
about high consumption patterns is desirable, both regarding group identification criteria (e.g. age or sex
differences, vegetarian or regional dietary customs, etc.) and statistical aspects.
CXS 193-1995 7
Dietary intake of contaminants: Reference is made to the Guidelines for the Study of Dietary Intake of
Chemical Contaminants.4 It is important to supply all relevant details, such as the type of study (duplicate diet,
total diet or market basket study, selective study), and statistical details. Calculated contaminant intake data
from food consumption models may also be useful. When results about food groups and effects of preparation
and cooking etc. are available, these should also be supplied.
Technological considerations
Information about the source of the contaminant and the way in which the food and feed is contaminated,
possibly including information, if available, about contamination being present in parts only of the product, is
essential for assessing the possibilities to control the contamination process and to be able to guarantee a
desired product safety and quality. Where possible source-related measures should be proposed. Good
management practices (GMP) and/or good agricultural practices (GAP) should also be adapted to control a
contamination problem. When this is possible, MLs may be based on GMP or GAP considerations to establish
at a level as low as reasonably achievable and necessary to protect the consumer. Considerations regarding
the technological possibilities to control a contamination problem, e.g. by cleaning, should also be considered
when a primary risk assessment model (theoretical maximum daily intake) shows possible intakes exceeding
the toxicological reference value. In such a case the possibilities of lower contamination levels need further
careful examination. Then a detailed study about all the aspects involved is necessary, so that decisions about
MLs can be based on a thorough evaluation of both the public health arguments and the potential problem
with complying with the proposed standard.
Risk assessment and risk management considerations
Risk assessment and risk management are conducted in accordance with the Working principles for risk
analysis for food safety for application by governments.
Establishment of maximum levels
In case it is decided that, on the basis of the outcome of the risk assessment, there is no need to establish an
ML to protect public health as the level of hazard/risk does not pose a public health problem, this should be
communicated in a transparent and accessible manner (e.g. by using the full format as provided for Schedule
I and to mention in the box of maximum level “not necessary”).
The establishment of MLs of contaminants in food and feed involves several principles, some of which
have already been mentioned in this preamble. Briefly stated, the following criteria will help in maintaining a
consistent policy in this matter:
 MLs should be set only for those contaminants that present both a significant risk to public health
and a known or expected problem in international trade.
 MLs should be set only for food that is significant for the total exposure of the consumer to the
contaminant. When identifying the significance of certain foods in the total exposure to the
contaminant, the criteria contained in the “Policy of the Codex Committee on Contaminants in
Foods for exposure assessment of contaminants and toxins in foods or food groups” (Section 4.5
of the Codex Procedural Manual) should be consulted.
 MLs should be set as low as reasonably achievable and at levels necessary to protect the
consumer. Providing it is acceptable from the toxicological point of view, MLs should be set at a
level which is (slightly) higher than the normal range of variation in levels in food and feed that are
produced with current adequate technological methods, in order to avoid undue disruptions of food
and feed production and trade. Where possible, MLs should be based on GMP and/or GAP
considerations in which the health concerns have been incorporated as a guiding principle to
achieve contaminant levels as low as reasonably achievable and necessary to protect the
consumer. Foods that are evidently contaminated by local situations or processing conditions that
can be avoided by reasonably achievable means shall be excluded in this evaluation unless a
higher ML can be shown to be acceptable from a public health point of view and significant
economic aspects are at stake.
 Proposals for MLs in products should be based on data from various countries and sources,
encompassing the main production areas/processes of those products, as far as they are engaged
in international trade. When there is evidence that contamination patterns are sufficiently
understood and will be comparable on a global scale, more limited data may be enough.
 MLs may be set for product groups when sufficient information is available about the contamination
pattern for the whole group, or when there are other arguments that extrapolation is appropriate.
 Numerical values for MLs should preferably be regular figures in a geometric scale (0.01, 0.02,
0.05, 0.1, 0.2, 0.5, 1, 2, 5, etc.), unless this may pose problems in the acceptability of the MLs.
CXS 193-1995 8
 MLs should apply to representative samples per lot. If necessary, appropriate methods of sampling
should be specified.
 MLs should not be lower than a level which can be analysed with methods of analysis that can
readily be set up and applied in food and feed control laboratories unless public health
considerations necessitate a lower ML which can only be controlled by means of a more elaborate
and sensitive method of analysis with an adequate lower detection limit. In all cases, a validated
method of analysis should be available with which a ML can be controlled.
 The contaminant as it should be analysed and to which the ML applies should be clearly defined.
The definition may include important metabolites when this is appropriate from an analytical or
toxicological point of view. It may also be aimed at indicator substances which are chosen from a
group of related contaminants.
 The product as it should be analysed and to which the ML applies, should be clearly defined. In
general, MLs are set on primary products. MLs should in general preferably be expressed as a
level of the contaminant related to the product as it is, on a fresh weight basis. In some cases,
however, there may be valid arguments to prefer expression on a dry weight basis (this might be
in particular the case for contaminants in feed) or on a fat weight basis (this might be in particular
the case for fat-soluble contaminants). Preferably the product should be defined as it moves in
trade, with provisions where necessary for the removal of inedible parts that might interfere with
the preparation and the analysis of the sample. The product definitions used by CCPR and
contained in the Classification of foods and animal feeds (CXA -1989)5 may serve as guidance on
this subject; other product definitions should only be used for specified reasons. For contaminant
purposes, however, analysis and consequently MLs should preferably be on the basis of the edible
part of the product.
 For fat-soluble contaminants, which may accumulate in animal products, provisions should be
applied regarding the application of the ML to products with various fat content (comparable to the
provisions for fat-soluble pesticides).
 Guidance is desirable regarding the possible application of MLs established for primary products
to processed products and multi-ingredient products. When products are concentrated, dried or
diluted, use of the concentration or dilution factor is generally appropriate in order to be able to
obtain a primary judgement of the contaminant levels in these processed products. The maximum
contaminant concentration in a multi-ingredient food and feed can likewise be calculated from the
composition of the food and feed. Information regarding the behaviour of the contaminant during
processing (e.g. washing, peeling, extraction, cooking, drying, etc.) is, however, desirable to give
more adequate guidance. When contaminant levels are consistently different in processed
products related to the primary products from which they are derived, and sufficient information is
available about the contamination pattern, it may be appropriate to establish separate MLs for
these processed products. This also applies when contamination may occur during processing. In
general, however, MLs should preferably be set for primary agricultural products and may be
applied to processed, derived and multi-ingredient food and feed by using appropriate conversion
factors. When these factors are sufficiently known, they should be mentioned in the suffix to the
ML following the format of list of MLs as defined in Annex II.
 MLs should preferably not be set higher than is acceptable in a primary (theoretical maximum
intake and risk estimation) approach of their acceptability from a public health point of view. When
this poses problems in relation to other criteria for establishing MLs, further evaluations are
necessary regarding the possibilities to reduce the contaminant levels, e.g. by improving GAP
and/or GMP conditions. When this does not bring a satisfactory solution, further refined risk
assessment and contaminant risk management evaluations will have to be made in order to try to
reach an agreement about an acceptable ML.
Procedure for risk assessment in relation to (proposed) MLs
It is more difficult to control food and feed contamination problems than in the case of food additives and
pesticide residues. Proposed MLs will inevitably be influenced by this situation. In order to promote acceptance
of Codex MLs, it is therefore important that assessments of the impact of those MLs on dietary exposure are
done in a consistent and realistic way. The procedure involves an assessment of the dietary intake in relation
to the proposed or existing MLs and the toxicological reference value.
In case a contaminant is carried over from feed to food of animal origin, the intake of a contaminant by the
different food-producing animal species and the resulting levels in the food of animal origin should be
estimated.
CXS 193-1995 9
The best estimate of dietary intake involves the national dietary pattern and corrections for concentration
changes during transport, storage, food preparation, for known levels in foods as consumed, etc. Caution is
recommended when using other than average food consumption values, although it is considered appropriate
to use relevant average food consumption data for identifiable subgroups of the population. Food consumption
patterns with a higher intake of critical foods may be used in the intake calculations when this is part of an
accepted national or international health protection and risk management policy. A harmonized approach using
an appropriate intake estimation model that is as realistic as possible is recommended. (cf. the “Policy of the
Codex Committee on Contaminants in Foods for exposure assessment of contaminants and toxins in foods or
food groups”,Section 4.5 of the Codex Procedural Manual). Calculated data should, where possible, always
be compared with measured intake data. Proposals for MLs should be accompanied by intake calculations
and risk assessment conclusions regarding their impact on dietary intake and use. The intake calculations
should follow the methodology described in the “Policy of the Codex Committee on Contaminants in Foods for
exposure assessment of contaminants and toxins in foods or food groups” and, if appropriate, be accompanied
by the generation of distribution curves for the concentration in specific foods/food groups (see Section 4.5 of
the Codex Procedural Manual). Statements from governments about the non-acceptance of (proposed) Codex
MLs should refer to specified intake calculations and risk management conclusions, which support this
position.
CXS 193-1995 10
Annex II
FORMAT OF THE GENERAL STANDARD FOR CONTAMINANTS AND TOXINS IN FOOD AND FEED
Introduction
The format for the schedule shall contain the following elements:
 Name of the contaminant.
 Synonyms: symbols, synonyms, abbreviations, scientific descriptions shall be mentioned.
 Reference to Joint FAO/WHO Expert Committee on Food Additives (JECFA) meetings (in
which the contaminant was discussed).
 Provisional maximum tolerable daily intake (PMTDI), provisional tolerable weekly intake
(PTWI) or similar toxicological guidance value: when the situation is complex, a short
statement and further references may be necessary here.
 Contaminant definition: definition of the contaminant as it shall be analysed and to which the
maximum level or guideline level applies.
 Reference to a source-directed measure or a related code of practice for the contaminant, if
appropriate.
 List of Codex maximum levels or guideline levels for that contaminant; this list shall be
composed of the following elements, in columns:
- feed/food commodity/product name;
- numerical value of maximum level or guideline level and units in which it is expressed;
- portion of the commodity/product to which the maximum level or guideline level applies;
and
- notes/remarks, including reference to relevant Codex commodity standards and where
necessary, definition of the commodity product.
CXS 193-1995 11
SCHEDULE
MAXIMUM AND GUIDELINE LEVELS FOR CONTAMINANTS AND TOXINS IN FOODS
INDEX OF CONTAMINANTS

NAME PAGE

Mycotoxins

Aflatoxins, total 14

Aflatoxin M1 42

Deoxynivalenol (DON) 43

Fumonisins 48

Ochratoxin A 53

Patulin 54

Metals

Arsenic 55

Cadmium 57

Lead 60

Mercury 66

Methylmercury 67

Tin 74

Radionuclides 75

Others

Acrylonitrile 80

Chloropropanols 81

Hydrocyanic acid 82

Melamine 83

Vinylchloride monomer 84
CXS 193-1995 12
EXPLANATORY NOTES

Reference to JECFA References to the Joint FAO/WHO Expert Committee on Food Additives (JECFA)
meeting in which the contaminant was evaluated and the year of that meeting.

Toxicological Toxicological advice about the tolerable intake level of the contaminant for
guidance value humans, expressed per kg body weight (bw). The year of recommendations and
additional explanation are included.

Contaminant Definition of the contaminant in the form of which the maximum level (ML) or
definition guideline level (GL) applies or which may or should be analysed in
commodities/products.

Synonyms Symbols, synonyms abbreviations, scientific descriptions and identification codes

used to define the contaminant.

Commodity/ The commodities or products to which the ML or GL applies, other than the terms
product name feed or food, are those that are intended for human consumption unless otherwise
specified.
The ML or GL contained in Codex commodity standards apply to the commodities
within the scope of the Codex commodity standard. Reference to the Codex
standard is provided and the definition of the commodity/product is the definition
as provided in the Codex commodity standard.
When the ML or GL applies only to the commodity within the scope of the Codex
commodity standard then the reference is mentioned as “Relevant Codex
commodity standard(s) is (are) …”. In case the reference to Codex commodity
standards is provided as an example for commodities to which the ML or GL
applies, then the reference is mentioned as “Relevant Codex commodity
standards include …”
For the other commodities or products not contained in Codex commodity
standards the definition of the commodity or product is provided in the
Classification of foods and animal feeds , unless otherwise specified.
In case a ML or GL applies to a product group (e.g. legume vegetables), the ML or
GL applies to all individual products belonging to the group as defined in the
classification.
For any other commodities or products other than those described above, where
necessary, the definition of the commodity/product is provided in “Notes/Remarks”.

Portion of the The portion of the feed or food to which the ML or GL applies is the portion defined
commodity/product in the Codex commodity standard or in the Classification of foods and animal
to which the feeds or defined at the establishment of the ML or GL, unless otherwise specified.
maximum level (ML)
or guideline level
(GL) applies
CXS 193-1995 13
DEFINITIONS OF SOME TOXICOLOGICAL TERMS

PMTDI Provisional maximum tolerable daily intake

The end-point used for contaminants with no cumulative properties. Its value
represents permissible human exposure as a result of the natural occurrence of
the substance in food and in drinking water. In the case of trace elements that are
both essential nutrients and unavoidable constituents of food, a range is
expressed, the lower value representing the level of essentiality and the upper
value the PMTDI.

PTWI Provisional tolerable weekly intake

An end-point used for food contaminants such as heavy metals with cumulative
properties. Its value represents permissible human weekly exposure to those
contaminants unavoidably associated with the consumption of otherwise
wholesome and nutritious foods.

PTMI Provisional tolerable monthly intake

An end-point used for a food contaminant with cumulative properties that has a
very long half-life in the human body. Its value represents permissible human
monthly exposure to a contaminant unavoidably associated with otherwise
wholesome and nutritious foods.
CXS 193-1995 14
AFLATOXINS, TOTAL
Reference to JECFA: 31 (1987), 46 (1996), 49 (1997), 68 (2007)
Toxicological guidance value: Carcinogenic potency estimates for aflatoxins B, G, M (1997, Intake should be reduced to levels as low as
reasonably possible)
Contaminant definition: Aflatoxins total (B1 + B2 + G1 + G2)
Synonyms: Abbreviations, AFB, AFG, with numbers, to designate specific compounds
Related code of practice: Code of practice for the prevention and reduction of aflatoxin contamination in peanuts (CXC 55-2004)6
Code of practice for the prevention and reduction of aflatoxin contamination in tree nuts (CXC 59-2005)7
Code of practice for the reduction of aflatoxin B1 in raw materials and supplemental feeding stuffs for milk producing
animals (CXC 45-1997)8
Code of practice for the prevention and reduction of aflatoxin contamination in dried figs (CXC 65-2008)9
Code of practice for the prevention and reduction of mycotoxins in spices (CXC 78-2017)10
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/remarks
product name µg/kg product to which the ML applies
Whole commodity after removal of The ML applies to almonds “ready-to-eat”.*
Almonds 10
shell. For sampling plan, see Annex IV.
Whole commodity after removal of The ML applies to almonds intended for further processing.**
Almonds 15
shell. For sampling plan, see Annex IV.
The ML applies to shelled Brazil nuts ready-to-eat.*
Brazil nuts 10 Whole commodity
For sampling plan, see Annex IV.
The ML applies to shelled Brazil nuts intended for further processing.**
Brazil nuts 15 Whole commodity
For sampling plan, see Annex IV.
Whole commodity after removal of The ML applies to hazelnuts, also known as filberts, “ready-to-eat”.*
Hazelnuts 10
shell. For sampling plan, see Annex IV.
The ML applies to hazelnuts, also known as filberts, intended for
Whole commodity after removal of further processing.**
Hazelnuts 15
shell.
For sampling plan, see Annex IV.
The ML applies for peanuts, also known as groundnuts, intended for
Unless specified, seed or kernels, after further processing.**
Peanuts 15
removal of shell or husk.
For sampling plan, see Annex III.
The ML applies to pistachios “ready-to-eat”.*
Whole commodity after removal of
Pistachios 10 For sampling plan, see Annex IV.
shell.
CXS 193-1995 15

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/remarks
product name µg/kg product to which the ML applies
Whole commodity after removal of The ML applies to pistachios intended for further processing.**
Pistachios 15
shell. For sampling plan, see Annex IV.
The ML applies to dried figs “ready-to-eat”.*
Dried figs 10 Whole commodity
For sampling plan, see Annex V.
The relevant Codex commodity standards are CXS 352-2022 and
Chili pepper, nutmeg 20 Whole/Powder/Crushed/Ground CXS 353-2022.
The ML applies to spices (dried/dry).
“Destined for further processing” means intended to undergo an
additional processing/treatment that has proven to reduce levels of
AFs before being used as an ingredient in foodstuffs, otherwise
Maize grain, destined processed or offered for human consumption. Codex members may
15 Whole commodity
for further processing define the processes that have been shown to reduce levels. The ML
does not apply to maize destined to animal feed or wet milling.
For sampling plan, see Annex VI.
Flour, meal, semolina
and flakes derived 10 Whole commodity For sampling plan, see Annex VI.
from maize

Husked rice 20 Whole commodity For sampling plan, see Annex VI.

Polished rice 5 Whole commodity For sampling plan, see Annex VI.
“Destined for further processing” means intended to undergo an
additional processing/treatment that has proven to reduce levels of
Sorghum grain, AFs before being used as an ingredient in foodstuffs, otherwise
destined for further 10 Whole commodity processed or offered for human consumption. Codex members may
processing define the processes that have been shown to reduce levels.
For sampling plan, see Annex VI.
The relevant Codex commodity standard is CXS 74-1981.
Cereal-based foods Whole commodity as sold; not
The ML applies to all cereal-based foods intended for infants
for infants and young 5 reconstituted or otherwise prepared for
(up to 12 months) and young children (12 to 36 months).
children consumption.
For sampling plan, see Annex VI.
CXS 193-1995 16

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/remarks
product name µg/kg product to which the ML applies

The relevant Codex commodity standard is CXS 74-1981.

Cereal-based foods Whole commodity as sold; not
for infants and young 10 reconstituted or otherwise prepared for The ML applies to cereal-based foods destined for food aid
children consumption. programmes intended for infants (6 to 12 months) and young children
(12 to 36 months).

Notes:
*“ready-to-eat” means not intended to undergo an additional processing/treatment that has proven to reduce levels of aflatoxins before being used as ingredient in
foodstuffs, otherwise processed or offered for human consumption
**“destined for further processing” means intended to undergo an additional processing/treatment that has proven to reduce levels of aflatoxins before being used
as an ingredient in foodstuffs, otherwise processed or offered for human consumption. Processes that have proven to reduce levels of aflatoxins are shelling,
blanching followed by colour sorting, and sorting by specific gravity and colour (damage). There is some evidence that roasting reduces aflatoxins in pistachios
but for other nuts the evidence is still to be supplied..
CXS 193-1995 17
Annex III
SAMPLING PLAN FOR TOTAL AFLATOXINS IN PEANUTS INTENDED FOR FURTHER PROCESSING
INTRODUCTION
1. The sampling plan calls for a single 20 kg laboratory sample of shelled peanuts (27 kg of unshelled
peanuts) to be taken from a peanut lot (sublot) and tested against a maximum level (ML) of 15 g/kg
total aflatoxins.
2. This sampling plan has been designed for enforcement and controls concerning total aflatoxins in bulk
consignments of peanuts traded in the export market. To assist Member Countries in implementing the
sampling plan, sample selection methods, sample preparation methods and analytical methods
required, to quantify aflatoxin in bulk peanut lots are described in this document.
A. DEFINITIONS

An identifiable quantity of a food commodity delivered at one time and determined by

Lot the official to have common characteristics, such as origin, variety, type of packing,
packer, consignor or markings.

Designated part of a large lot in order to apply the sampling method on that designated
Sublot
part. Each sublot must be physically separate and identifiable.

It is defined by an aflatoxin test procedure and an accept/reject limit. An aflatoxin test

procedure consists of three steps: sample selection, sample preparation and aflatoxin
Sampling plan
quantification. The accept/reject limit is a tolerance usually equal to the Codex
maximum level.

Incremental
A quantity of material taken from a single random place in the lot or sublot.
sample

The combined total of all the incremental samples taken from the lot or sublot. The
Aggregate sample
aggregate sample has to be at least as large as the 20 kg laboratory sample.

The smallest quantity of peanuts comminuted in a mill. The laboratory sample may be
a portion of or the entire aggregate sample. If the aggregate sample is larger than
Laboratory sample 20 kg, a 20 kg laboratory sample should be removed in a random manner from the
aggregate sample. The sample should be finely ground and mixed thoroughly using a
process that approaches as complete a homogenization as possible.

A portion of the comminuted laboratory sample. The entire 20 kg laboratory sample

should be comminuted in a mill. A portion of the comminuted 20 kg sample is randomly
Test portion removed for the extraction of the aflatoxin for chemical analysis. Based upon grinder
capacity, the 20 kg aggregate sample can be divided into several equal-sized samples,
if all results are averaged.

B. SAMPLING
Material to be sampled
3. Each lot, which is to be examined, must be sampled separately. Large lots should be subdivided into sublots
to be sampled separately. The subdivision can be done following provisions laid down in Table 1 below.
4. Considering that the weight of the lot is not always an exact multiple of the weight of the sublots, the weight
of the sublot may exceed the mentioned weight by a maximum of 20 percent.
Table 1. Subdivision of large lots into sublots for sampling
Number of Laboratory
Lot weight – Weight or
Commodity incremental sample
tonne (T) number of sublots
samples weight (kg)
 500 100 tonnes 100 20
> 100 and < 500 5 sublots 100 20
Peanuts
 25 and  100 25 tones 100 20
> 15 and <= 25 --1 sublot 100 20
CXS 193-1995 18
Number of incremental samples for lots of less than 15 tonnes
5. The number of incremental samples to be taken depends on the weight of the lot, with a minimum of ten
and a maximum of 100. The figures in the following Table 2 may be used to determine the number of
incremental samples to be taken. It is necessary that the total sample weight of 20 kg is achieved.
Table 2. Number of incremental samples to be taken depending on the weight of the lot

Lot weight tones – (T) N° of incremental samples

T1 10
1<T5 40
5 < T  10 60
10 < T < 15 80
Incremental sample selection
6. Procedures used to take incremental samples from a peanut lot are extremely important. Every
individual peanut in the lot should have an equal chance of being chosen. Biases will be introduced by
the sample selection methods if equipment and procedures used to select the incremental samples
prohibit or reduce the chances of any item in the lot from being chosen.
7. Since there is no way to know if the contaminated peanut kernels are uniformly dispersed throughout
the lot, it is essential that the aggregate sample be the accumulation of many small portions or
increments of the product selected from different locations throughout the lot. If the aggregate sample
is larger than desired, it should be blended and subdivided until the desired laboratory sample size is
achieved.
Static lots
8. A static lot can be defined as a large mass of peanuts contained either in a single large container such
as a wagon, truck, or railcar or in many small containers such as sacks or boxes and the peanuts are
stationary at the time a sample is selected. Selecting a truly random sample from a static lot can be
difficult because the container may not allow access to all peanuts.
9. Taking an aggregate sample from a static lot usually requires the use of probing devices to select
product from the lot. The probing devices used should be specially designed for the type of container.
The probe should (1) be long enough to reach all product, (2) not restrict any item in the lot from being
selected, and (3) not alter the items in the lot. As mentioned above, the aggregate sample should be a
composite from many small increments of product taken from many different locations throughout the
lot.
10. For lots traded in individual packages, the sampling frequency (SF), or number of packages that
incremental samples are taken from, is a function of the lot weight (LT), incremental sample weight (IS),
aggregate sample weight (AS) and the individual packing weight (IP), as follows:
Equation 1: SF = (LT x IS) / (AS x IP)
The SF is the number of packages sampled. All weights should be in the same mass units such as kg.
Dynamic lots
11. True random sampling can be more nearly achieved when selecting an aggregate sample from a moving
stream of peanuts as, the lot is transferred, for example, by a conveyor belt from one location to another.
When sampling from a moving stream, take small increments of product from the entire length of the
moving stream; composite the peanuts to obtain an aggregate sample; if the aggregate sample is larger
than the required laboratory sample, then blend and subdivide the aggregate sample to obtain the
desired size laboratory sample.
12. Automatic sampling equipment such as cross-cut samplers are commercially available with timers that
automatically pass a diverter cup through the moving stream at predetermined and uniform intervals.
When automatic equipment is not available, a person can be assigned to manually pass a cup though
the stream at periodic intervals to collect incremental samples. Whether using automatic or manual
methods, small increments of peanuts should be collected and composited at frequent and uniform
intervals throughout the entire time peanuts flow past the sampling point.
CXS 193-1995 19
13. Cross-cut samplers should be installed in the following manner: (1) the plane of the opening of the
diverter cup should be perpendicular to the direction of flow; (2) the diverter cup should pass through
the entire cross-sectional area of the stream; and (3) the opening of the diverter cup should be wide
enough to accept all items of interest in the lot. As a general rule, the width of the diverter cup opening
should be about three times the largest dimensions of the items in the lot.
14. The size of the aggregate sample (S) in kg, taken from a lot by a cross-cut sampler is:
Equation 2: S = (D x LT) / (T x V)
D is the width of the diverter cup opening (in cm), LT is the lot size (in kg), T is interval or time between
cup movement through the stream (in seconds), and V is cup velocity (in cm/sec).
15. If the mass flow rate of the moving stream, MR (kg/sec), is known, then the sampling frequency (SF), or
number of cuts made by the automatic sampler cup is:
Equation 3: SF = (S x V) / (D x MR)
16. Equation 2 can also be used to compute other terms of interest such as the time between cuts (T). For
example, the required time (T) between cuts of the diverter cup to obtain a 20 kg aggregate sample from
a 30 000 kg lot where the diverter cup width is 5.08 cm (2 inches), and the cup velocity through the
stream 30 cm/sec. Solving for T in Equation 2.
T = (5.08 cm x 30 000 kg) / (20 kg x 30 cm/sec) = 254 sec
17. If the lot is moving at 500 kg per minute, the entire lot will pass through the sampler in 60 minutes and
only 14 cuts (14 incremental samples) will be made by the cup through the lot. This may be considered
too infrequent in that too much product passes through the sampler between the time the cup cuts
through the stream.
Weight of the incremental sample
18. The weight of the incremental sample should be approximately 200 g or greater, depending on the total
number of increments, to obtain an aggregate sample of 20 kg.
Packaging and transmission of samples
19. Each laboratory sample shall be placed in a clean, inert container offering adequate protection from
contamination and against damage in transit. All necessary precautions shall be taken to avoid any change
in composition of the laboratory sample which might arise during transportation or storage.
Sealing and labelling of samples
20. Each laboratory sample taken for official use shall be sealed at the place of sampling and identified. A record
must be kept of each sampling, permitting each lot to be identified unambiguously and giving the date and
place of sampling together with any additional information likely to be of assistance to the analyst.
C. SAMPLE PREPARATION
Precautions
21. Daylight should be excluded as much as possible during the procedure since aflatoxin gradually breaks
down under the influence of ultraviolet light.
Homogenization – Grinding
22. As the distribution of aflatoxin is extremely non-homogeneous, samples should be prepared – and
especially homogenized – with extreme care. All laboratory sample obtained from aggregate sample is
to be used for the homogenization/grinding of the sample.
23. The sample should be finely ground and mixed thoroughly using a process that approaches as complete
a homogenization as possible.
24. The use of a hammer mill with a #14 screen (3.1 mm diameter hole in the screen) has been proven to
represent a compromise in terms of cost and precision. A better homogenization (finer grind – slurry)
can be obtained by more sophisticated equipment, resulting in a lower sample preparation variance.
Test portion
25. A minimum test portion size of 100 g taken from the laboratory sample.
CXS 193-1995 20
D. ANALYTICAL METHODS
Background
26. A criteria-based approach, whereby a set of performance criteria is established with which the analytical
method used should comply, is appropriate. The criteria-based approach has the advantage that, by
avoiding setting down specific details of the method used, developments in methodology can be
exploited without having to reconsider or modify the specified method. The performance criteria
established for methods should include all the parameters that need to be addressed by each laboratory,
such as the detection limit, repeatability coefficient of variation, reproducibility coefficient of variation and
the percent recovery necessary for various statutory limits. Utilizing this approach, laboratories would
be free to use the analytical method most appropriate for their facilities. Analytical methods that are
accepted by chemists internationally such as the Association of Official Analytical Collaboration (AOAC
International) may be used. These methods are regularly monitored and improved depending upon
technology.
Performance criteria for methods of analysis
Table 3. Specific requirements with which methods of analysis should comply

Maximum permitted
Criterion Concentration range Recommended value
value
Blanks All Negligible -
Recovery-Aflatoxins Total 1 – 15 g/kg 70 to 110%
> 15 g/kg 80 to 110%
Precision RSDR All As derived from 2 x value derived from
Horwitz equation Horwitz equation
Precision RSDr may be calculated as 0.66 times Precision RSDR at the concentration of interest.

 The detection limits of the methods used are not stated, as the precision values are given at the
concentrations of interest.
 The precision values are calculated from the Horwitz equation, i.e.:
(1 - 0.5logC)
RSD =2
R
where:
 RSDR is the relative standard deviation calculated from results generated under reproducibility
conditions [(Sr / x ) x 100]
 C is the concentration ratio (i.e. 1 = 100 g/100 g, 0.001 = 1 000 mg/kg)
27. This is a generalized precision equation, which has been found to be independent of analyte and matrix
but solely dependent on concentration for most routine methods of analysis.
CXS 193-1995 21
Annex IV
SAMPLING PLANS FOR AFLATOXIN CONTAMINATION IN
READY-TO-EAT TREE NUTS AND TREE NUTS
DESTINED FOR FURTHER PROCESSING: ALMONDS, HAZELNUTS, PISTACHIOS
AND SHELLED BRAZIL NUTS
DEFINITIONS

An identifiable quantity of a food commodity delivered at one time and

Lot determined by the official to have common characteristics, such as origin,
variety, type of packing, packer, consignor or markings.

Designated part of a larger lot in order to apply the sampling method on that
Sublot
designated part. Each sublot must be physically separate and identifiable.

It is defined by an aflatoxin test procedure and an accept/reject limit. An

aflatoxin test procedure consists of three steps: sample selection, sample
Sampling plan
preparation and aflatoxin quantification. The accept/reject limit is a
tolerance usually equal to the Codex maximum level (ML).

The quantity of material taken from a single random place in the lot or
Incremental sample
sublot.

The combined total of all the incremental samples that is taken from the lot
Aggregate sample or sublot. The aggregate sample has to be at least as large as the
laboratory sample or samples combined.

The smallest quantity of tree nuts comminuted in a mill. The laboratory

sample may be a portion of or the entire aggregate sample. If the aggregate
Laboratory sample
sample is larger than the laboratory sample(s), the laboratory sample(s)
should be removed in a random manner from the aggregate sample.

A portion of the comminuted laboratory sample. The entire laboratory

sample should be comminuted in a mill. A portion of the comminuted
Test portion
laboratory sample is randomly removed for the extraction of the aflatoxin for
chemical analysis.

Nuts, which are not intended to undergo an additional processing/treatment

that has proven to reduce levels of aflatoxins before being used as an
Ready-to-eat tree nuts
ingredient in foodstuffs, otherwise processed or offered for human
consumption.

Nuts, which are intended to undergo an additional processing/treatment that

has proven to reduce levels of aflatoxins before being used as an ingredient
in foodstuffs, otherwise processed or offered for human consumption.
Tree nuts destined for
Processes that have proven to reduce levels of aflatoxins are shelling,
further processing
blanching followed by colour sorting, and sorting by specific gravity and
colour (damage). There is some evidence that roasting reduces aflatoxins in
pistachios but for other nuts the evidence is still to be supplied.

A plot of the probability of a accepting a lot versus lot concentration when

Operating
using a specific sampling plan design. The OC curve provides an estimate
characteristic
of good lots rejected (exporter’s risk) and bad lots accepted (importer’s risk)
(OC) curve
by a specific aflatoxin sampling plan design.

SAMPLING PLAN DESIGN CONSIDERATIONS

1. Importers may commercially classify tree nuts as either “ready-to-eat” (RTE) or “destined for further
processing” (DFP). As a result, MLs and sampling plans are proposed for both commercial types of tree
nuts. MLs need to be defined for tree nuts DFP and RTE tree nuts before a final decision can be made
about a sampling plan design.
CXS 193-1995 22
2. Tree nuts can be marketed either as in-shell or shelled nuts. For example, pistachios are predominately
marketed as in-shell nuts while almonds are predominately marketed as shelled nuts.
3. Sampling statistics, shown in Annex IV.A, are based upon the uncertainty and aflatoxin distribution
among laboratory samples of shelled nuts. Because the shelled nut count per kg is different for each of
the tree nuts, the laboratory sample size is expressed in number of nuts for statistical purposes.
However, the shelled nut count per kg for each tree nut, shown in Annex IV.A, can be used to convert
laboratory sample size from number of nuts to mass and vice versa.
4. Uncertainty estimates associated with sampling, sample preparation and analysis, shown in Annex IV.A,
and the negative binomial distribution are used to calculate OC curves that describe the performance of
the proposed aflatoxin sampling plans.
5. In Annex IV.A, the analytical variance reflects a reproducibility relative standard deviation of 22 percent,
which is based upon Food Analysis Performance Assessment Scheme (FAPAS) data. A relative
standard deviation of 22 percent is considered by FAPAS as an appropriate measure of the best
agreement that can be reliably obtained between laboratories. An analytical uncertainty of 22 percent is
larger than the within laboratory variation measured in the sampling studies for the four tree nuts.
6. The issue of correcting the analytical test result for recovery is not addressed in this document. However,
Table 2 specifies several performance criteria for analytical methods including suggestions for the range
of acceptable recovery rates.
AFLATOXIN TEST PROCEDURE AND MAXIMUM LEVELS
7. An aflatoxin sampling plan is defined by an aflatoxin test procedure and a maximum level. A value for
the ML and the aflatoxin test procedure are given below in this section.
8. The MLs for total aflatoxins in tree nuts (almonds, hazelnuts, pistachios, and shelled Brazil nuts) “RTE”
and “destined for further processing” are 10 µg/kg and 15 µg/kg, respectively.
9. Choice of the number and size of the laboratory sample is a compromise between minimizing risks (false
positives and false negatives) and costs related to sampling and restricting trade. For simplicity, it is
recommended that the proposed aflatoxin sampling plans use a 20 kg aggregate sample for all four tree
nuts.
10. The two sampling plans (RTE and DFP) have been designed for enforcement and controls concerning
total aflatoxins in bulk consignments (lots) of tree nuts traded in the export market.
Tree nuts destined for further processing
Maximum level – 15 µg/kg total aflatoxins
Number of laboratory samples – 1
Laboratory sample size – 20 kg
Almonds – shelled nuts
Hazelnuts – shelled nuts
Pistachios – in-shell nuts (equivalent to about 10 kg shelled nuts that is calculated on the
basis of the actual edible portion in the sample)
Brazil nuts – shelled nuts
Sample preparation – sample shall be finely ground and mixed thoroughly using a
process, e.g. dry grind with a vertical cutter mixer type mill, that
has been demonstrated to provide the lowest sample preparation
variance. Preferably, Brazil nuts should be ground as slurry.
Analytical method – performance based (see Table 2)
Decision rule – If the aflatoxin test result is less than or equal to 15 µg/kg total
aflatoxins, then accept the lot. Otherwise, reject the lot.
Ready-to-eat tree nuts
Maximum level – 10 µg/kg total aflatoxins
Number of laboratory samples – 2
CXS 193-1995 23
Laboratory sample size – 10 kg
Almonds – shelled nuts
Hazelnuts – shelled nuts
Pistachios – in-shell nuts (equivalent to about 5 kg shelled nuts per test sample that is
calculated on the basis of the actual edible portion in the sample)
Brazil nuts – shelled nuts
Sample preparation – sample shall be finely ground and mixed thoroughly using a
process, e.g. dry grind with a vertical cutter mixer type mill, that
has been demonstrated to provide the lowest sample preparation
variance. Preferably, Brazil nuts should be ground as slurry.
Analytical method – performance based (see Table 2)
Decision rule – if the aflatoxin test result is less than or equal to 10 µg/kg total
aflatoxin in both test samples, then accept the lot. Otherwise,
reject the lot.
11. To assist Member Countries implement these two sampling plans, sample selection methods, sample
preparation methods, and analytical methods required to quantify aflatoxin in laboratory samples taken
from bulk tree nut lots are described in the following sections.
SAMPLE SELECTION
M ATERIAL TO BE SAMPLED
12. Each lot, which is to be examined for aflatoxin, must be sampled separately. Lots larger than 25 tonnes
should be subdivided into sublots to be sampled separately. If a lot is greater than 25 tones, the number
of sublots is equal to the lot weight in tonnes divided by 25 tones. It is recommended that a lot or a sublot
should not exceed 25 tonnes. The minimum lot weight should be 500 kg.
13. Considering that the weight of the lot is not always an exact multiple of 25 tone sublots, the weight of
the sublot may exceed the mentioned weight by a maximum of 25 percent.
14. Samples should be taken from the same lot, i.e. they should have the same batch code or at the very
least the same best before date. Any changes, which would affect the mycotoxin content, the analytical
determination or make the aggregate samples collected unrepresentative should be avoided. For
example, do not open packaging in adverse weather conditions or expose samples to excessive
moisture or sunlight. Avoid cross-contamination from other potentially contaminated consignments
nearby.
15. In most cases any truck or container will have to be unloaded to allow representative sampling to be
carried out.
INCREMENTAL SAMPLE SELECTION
16. Procedures used to take incremental samples from a tree nut lot are extremely important. Every
individual nut in the lot should have an equal chance of being chosen. Biases will be introduced by
sample selection methods if equipment and procedures used to select the incremental samples prohibit
or reduce the chances of any item in the lot from being chosen.
17. Since there is no way to know if the contaminated tree nut kernels are uniformly dispersed throughout
the lot, it is essential that the aggregate sample be the accumulation of many small incremental samples
of product selected from different locations throughout the lot. If the aggregate sample is larger than
desired, it should be blended and subdivided until the desired laboratory sample size is achieved.
NUMBER OF INCREMENTAL SAMPLES FOR LOTS OF VARYING WEIGHT
18. The number and size of the laboratory sample(s) will not vary with lot (sublot) size. However, the number
and size of the incremental samples will vary with lot (sublot) size.
19. The number of incremental samples to be taken from a lot (sublot) depends on the weight of the lot.
Table 1 shall be used to determine the number of incremental samples to be taken from lots or sublots
of various sizes below 25 tonnes. The number of incremental samples varies from a minimum of 10 and
to a maximum of 100.
CXS 193-1995 24
Table 1. Number and size of incremental samples composited for an aggregate sample of 20 kga as a
function of lot (or sublot) weight

Minimum incremental Minimum aggregate

Lot or sublot weightb Minimum number of
sample sizec sample size
(T in tonnes) incremental samples
(g) (Kg)
T<1 10 2 000 20
1T<5 25 800 20
5  T < 10 50 400 20
10  T < 15 75 267 20
15 T 100 200 20

Notes
a Minimum aggregate sample size = laboratory sample size of 20 kg
b 1 tonne = 1 000 kg
c Minimum incremental sample size = laboratory sample size (20 kg)/minimum number of incremental
samples, i.e. for 0.5 < T < 1 tonne, 2 000 g = 20 000/10
WEIGHT OF THE INCREMENTAL SAMPLE
20. The suggested minimum weight of the incremental sample should be approximately 200 g for lots of
25 metric tonnes (MT) (25 000 kg). The number and/or size of incremental samples will have to be larger
than that suggested in Table 1 for lots sizes below 25 000 kg in order to obtain an aggregate sample
greater than or equal to the 20 kg laboratory sample.
STATIC LOTS
21. A static lot can be defined as a large mass of tree nuts contained either in a large single container such
as a wagon, truck or railcar or in many small containers such as sacks or boxes and the nuts are
stationary at the time a sample is selected. Selecting a truly random sample from a static lot can be
difficult because all containers in the lot or sublot may not be accessible.
22. Taking incremental samples from a static lot usually requires the use of probing devices to select product
from the lot. The probing devices should be specifically designed for the commodity and type of
container. The probe should: (1) be long enough to reach all products; (2) not restrict any item in the lot
from being selected; and (3) not alter the items in the lot. As mentioned above, the aggregate sample
should be a composite from many small incremental samples of product taken from many different
locations throughout the lot.
23. For lots traded in individual packages, the sampling frequency (SF), or number of packages that
incremental samples are taken from, is a function of the lot weight (LT), incremental sample weight (IS),
aggregate sample weight (AS) and the individual packing weight (IP), as follows:
Equation 1: SF = (LT x IS) / (AS x IP)
24. The SF is the number of packages sampled. All weights should be in the same mass units such as kg.
DYNAMIC LOTS
25. Representative aggregate samples can be more easily produced when selecting incremental samples
from a moving stream of tree nuts as the lot is transferred from one location to another. When sampling
from a moving stream, take small incremental samples of product from the entire length of the moving
stream; composite the incremental samples to obtain an aggregate sample; if the aggregate sample is
larger than the required laboratory sample(s), then blend and subdivide the aggregate sample to obtain
the desired size laboratory sample(s).
26. Automatic sampling equipment such as a cross-cut sampler is commercially available with timers that
automatically pass a diverter cup through the moving stream at predetermined and uniform intervals.
When automatic sampling equipment is not available, a person can be assigned to manually pass a cup
through the stream at periodic intervals to collect incremental samples. Whether using automatic or
manual methods, incremental samples should be collected and composited at frequent and uniform
intervals throughout the entire time the nuts flow past the sampling point.
CXS 193-1995 25
27. Cross-cut samplers should be installed in the following manner: (1) the plane of the opening of the
diverter cup should be perpendicular to the direction of the flow; (2) the diverter cup should pass through
the entire cross-sectional area of the stream; and (3) the opening of the diverter cup should be wide
enough to accept all items of interest in the lot. As a general rule, the width of the diverter cup opening
should be about two to three times the largest dimensions of items in the lot.
28. The size of the aggregate sample (S) in kg, taken from a lot by a cross-cut sampler is:
Equation 2: S = (D x LT) / (T x V)
where D is the width of the diverter cup opening (cm), LT is the lot size (kg), T is interval or time between
cup movement through the stream (seconds), and V is cup velocity (cm/sec).
29. If the mass flow rate of the moving stream, MR (kg/sec), is known, then the SF, or number of cuts made
by the automatic sampler cup, can be computed from Equation 3 as a function of S, V, D, and MR.
Equation 3: SF = (S x V) / (D x MR)
30. Equations 2 and 3 can also be used to compute other terms of interest such as the time between cuts
(T). For example, the time (T) required between cuts of the diverter cup to obtain a 20 kg aggregate
sample from a 20 000 kg lot where the diverter cup width is 5.0 cm and the cup velocity through the
stream 30 cm/sec. Solving for T in Equation 2.
T = (5.0 cm x 20 000 kg) / (20 kg x 20 cm/sec) = 250 sec
31 If the lot is moving at 500 kg per minute, the entire lot will pass through the sampler in 40 minutes
(2 400 sec) and only 9.6 cuts (9 incremental samples) will be made by the cup through the lot
(Equation 3). This may be considered too infrequent, in that too much product (2 083.3 kg) passes
through the sampler between the time the cup cuts through the stream.
PACKAGING AND TRANSPORTATION OF SAMPLES
32. Each laboratory sample shall be placed in a clean, inert container offering adequate protection from
contamination, sunlight and against damage in transit. All necessary precautions shall be taken to avoid
any change in composition of the laboratory sample, which might arise during transportation or storage.
Samples should be stored in a cool dark place.
SEALING AND LABELLING OF SAMPLES
33. Each laboratory sample taken for official use shall be sealed at the place of sampling and identified. A
record must be kept of each sampling, permitting each lot to be identified unambiguously and giving the
date and place of sampling together with any additional information likely to be of assistance to the
analyst.
SAMPLE PREPARATION
PRECAUTIONS
34. Sunlight should be excluded as much as possible during sample preparation, since aflatoxin gradually
breaks down under the influence of ultraviolet light. Also, environmental temperature and relative
humidity should be controlled and not favour mould growth and aflatoxin formation.
HOMOGENIZATION – GRINDING
35. As the distribution of aflatoxin is extremely non-homogeneous, laboratory samples should be
homogenized by grinding the entire laboratory sample received by the laboratory. Homogenization is a
procedure that reduces particle size and disperses the contaminated particles evenly throughout the
comminuted laboratory sample.
36. The laboratory sample should be finely ground and mixed thoroughly using a process that approaches
as complete homogenization as possible. Complete homogenization implies that particle size is
extremely small, and the variability associated with sample preparation (Annex IV.A) approaches zero.
After grinding, the grinder should be cleaned to prevent aflatoxin cross-contamination.
37. The use of vertical cutter mixer type grinders that mix and comminute the laboratory sample into a paste
represent a compromise in terms of cost and fineness of grind or particle size reduction. A better
homogenization (finer grind), such as a liquid slurry, can be obtained by more sophisticated equipment
and should provide the lowest sample preparation variance.
TEST PORTION
38. The suggested weight of the test portion taken from the comminuted laboratory sample should be
approximately 50 g. If the laboratory sample is prepared using a liquid slurry, the slurry should contain
50 g of nut mass.
CXS 193-1995 26
39. Procedures for selecting the 50 g test portion from the comminuted laboratory sample should be a
random process. If mixing occurred during or after the comminution process, the 50 g test portion can
be selected from any location throughout the comminuted laboratory sample. Otherwise, the 50 g test
portion should be the accumulation of several small portions selected throughout the laboratory sample.
40. It is suggested that three test portions be selected from each comminuted laboratory sample. The three
test portions will be used for enforcement, appeal and confirmation if needed.
ANALYTICAL METHODS
BACKGROUND
41. A criteria-based approach, whereby a set of performance criteria is established with which the analytical
method used should comply, is appropriate. The criteria-based approach has the advantage that, by
avoiding setting down specific details of the method used, developments in methodology can be
exploited without having to reconsider or modify the specific method. The performance criteria
established for methods should include all the parameters that need to be addressed by each laboratory
such as the detection limit, repeatability coefficient of variation (within lab), reproducibility coefficient of
variation (among lab), and the percent recovery necessary for various statutory limits. Analytical
methods that are accepted by chemists internationally such as the Association of Official Analytical
Collaboration (AOAC International) and the International Organization for Standardization (ISO) may be
used. These methods are regularly monitored and improved depending upon technology.
PERFORMANCE CRITERIA FOR METHODS OF ANALYSIS
42. A list of criteria and performance levels are shown in Table 2. Utilizing this approach, laboratories would
be free to use the analytical method most appropriate for their facilities.
Table 2. Specific requirements with methods of analysis should comply with
Concentration Recommended
Criterion Maximum permitted value
range (ng/g) value
Blanks All Negligible n/a
1 to 15 70 to 100% n/a
Recovery
> 15 80 to 110% n/a
Precision or relative 1 to 120 Equation 4 2 x value derived from Equation 4
standard deviation
RSDR > 120 Equation 5 2 x value derived from Equation 5
(Reproducibility)
Calculated as
1 to 120 0.66 times n/a
Precision or relative Precision RSDR
standard deviation
RSDr (Repeatability) Calculated as
> 120 0.66 times n/a
Precision RSDr
n/a = not applicable
43. The detection limits of the methods used are not stated. Only the precision values are given at the
concentrations of interest. The precision values are calculated from equations 4 and 5.
Equation 4: RSDR = 22.0 (for C ≤ 120 µg/kg or c ≤ 120 x 10-9)
Equation 5: RSDR = 2 (1-0.5logc) (for C > 120 µg/kg or c > 120 x 10-9)
where:
 RSDR = the relative standard deviation calculated from results generated under
reproducibility conditions
 RSDr = the relative standard deviation calculated from results generated under
repeatability conditions = 0.66 RSDR
 c = the aflatoxin concentration ratio
(i.e. 1 = 100 g/100 g, 0.001 = 1 000 mg/kg)
 C = aflatoxin concentration or mass of aflatoxin to mass of tree nuts
(i.e. µg/kg)
44. Equations 4 and 5 are generalized precision equations, which have been found to be independent of
analyte and matrix but solely dependent on concentration for most routine methods of analysis.
45. Results should be reported on the edible portion of the sample.
 CXS 193-1995 27
Annex IV.A
Uncertainty, as measured by the variance, associated with sampling, sample preparation and
analytical steps of the aflatoxin test procedure used to estimate aflatoxin in almonds, hazelnuts,
pistachios and shelled Brazil nuts.
Sampling data for almonds, hazelnuts, pistachios and shelled Brazil nuts were supplied by the United States
of America, Türkiye, the Islamic Republic of Iran and Brazil, respectively.
Sampling, sample preparation and analytical variances associated with testing almonds, hazelnuts, pistachios,
and shelled Brazil nuts are shown in Table 1 below.
Table 1. Variancesa associated with the aflatoxin test procedure for each tree nut

Test procedure Almonds Hazelnuts Pistachios Shelled Brazil nuts

Samplingb,c S2s = (7 730/ns) 5.759C1.561 S2s = (10 000/ns) 4.291C1.609 S2s = 8 000/ns) 7.913C1.475 ss2 = (1 850/ns) 4.8616C1.889

Sample
S2sp = (100/nss) 0.170C1.646 S2sp = (50/nss) 0.021C1.545 S2sp = (25/nss) 2.334C1.522 sss2 = (50/nss) 0.0306C0.632
Preparationd

experimental
sa2 = (1/n) 0.0164C1.117
Analyticale S2a = (1/na) 0.0484C2.0 S2a = (1/na) 0.0484C2.0 S2a = (1/na) 0.0484C2.0 or
FAPAS
sa2 = (1/n) 0.0484C2.0

Total variance S2s + S2sp + S2a S2s + S2sp + S2a S2s + S2sp + S2a S2s + S2sp + S2a

Notes
aVariance = S2 (s, sp, and a denote sampling, sample preparation and analytical steps, respectively, of aflatoxin test
procedure)
bns = laboratory sample size in number of shelled nuts, nss = test portion size in grams, na = number of aliquots quantified
by HPLC, and C = aflatoxin concentration in µg/kg total aflatoxin.
c Shelled nut count/kg for almonds, hazelnuts, pistachios, and Brazil nuts is 773, 1 000, 1 600 and 185, respectively.
d
Sample preparation for almonds, hazelnuts and pistachios reflects Hobart, Robot Coupe, Marjaan Khatman and Turrax
type mills, respectively. Laboratory samples were dry ground into a paste for each tree nut except for Brazil nut that were
prepared as a slurry Brazil nut/water 1/1 w/w.
e Analytical variances reflect FAPAS recommendation for upper limit of analytical reproducibility uncertainty. A relative
standard deviation of 22 percent, which is based upon FAPAS data, is considered, as an appropriate measure of the best
agreement that can be obtained between laboratories. An analytical uncertainty of 22 percent is larger than the within
laboratory uncertainty measured in the sampling studies for the four tree nuts.
CXS 193-1995 28
Annex V
SAMPLING PLAN FOR AFLATOXIN CONTAMINATION IN DRIED FIGS
DEFINITIONS

An identifiable quantity of a food commodity delivered at one time and

Lot determined by the official to have common characteristics, such as origin,
variety, type of packing, packer, consignor or markings.

Designated part of a larger lot in order to apply the sampling method on that
Sublot
designated part. Each sublot must be physically separate and identifiable.

It is defined by an aflatoxin test procedure and an accept/reject level. An

aflatoxin test procedure consists of three steps: sample selection of
Sampling plan sample(s) of a given size, sample preparation and aflatoxin quantification.
The accept/reject level is a tolerance usually equal to the Codex maximum
level (ML).

Incremental sample The quantity of material taken from a single random place in the lot or sublot.

The combined total of all the incremental samples that is taken from the lot or
Aggregate sample sublot. The aggregate sample has to be at least as large as the laboratory
sample or samples combined.

The smallest quantity of dried figs comminuted in a mill. The laboratory

sample may be a portion of or the entire aggregate sample. If the aggregate
Laboratory sample
sample is larger than the laboratory sample(s), the laboratory sample(s)
should be removed in a random manner from the aggregate sample.

A portion of the comminuted laboratory sample. The entire laboratory sample

should be comminuted in a mill. A portion of the comminuted laboratory
Test portion
sample is randomly removed for the extraction of the aflatoxin for chemical
analysis.

Dried figs, which are not intended to undergo an additional

Ready-to-eat processing/treatment that have proven to reduce levels of aflatoxin before
dried figs being used as an ingredient in foodstuffs, otherwise processed or offered for
human consumption.

A plot of the probability of accepting a lot versus lot concentration when using
Operating
a specific sampling plan design. The OC curve also provides an estimate of
characteristic
good lots rejected (exporter’s risk) and bad lots accepted (importer’s risk) by
(OC) curve
a specific aflatoxin sampling plan design.

SAMPLING PLAN DESIGN CONSIDERATIONS

1. Importers commercially classify dried figs mostly as “ready-to-eat” (RTE). As a result, MLs and sampling
plans are established only for RTE dried figs.
2. The performance of the sampling plan was computed using the variability and aflatoxin distribution
among laboratory samples of dried figs taken from contaminated lots. Because the dried fig count per kg
is different for different varieties of dried figs, the laboratory sample size is expressed in number of dried
figs for statistical purposes. However, the dried fig count per kg for each variety of dried figs can be used
to convert laboratory sample size from number of dried figs to mass and vice versa.
3. Uncertainty estimates (variances) associated with sampling, sample preparation and analysis and the
negative binomial distribution are used to calculate OC curves that describe the performance of the
aflatoxin sampling plans for dried figs.
4. The analytical variance measured in the sampling study reflects within laboratory variance and was
replaced with an estimate of analytical variance reflects a reproducibility relative standard deviation of
22 percent, which is based upon Food Analysis Performance Assessment Scheme (FAPAS) data. A
relative standard deviation of 22 percent is considered by FAPAS as an appropriate measure of the best
agreement that can be reliably obtained between laboratories. An analytical uncertainty of 22 percent is
larger than the within laboratory variation measured in the sampling studies for dried figs.
CXS 193-1995 29
5. The issue of correcting the analytical test result for recovery is not addressed in this document. However,
Table 2 specifies several performance criteria for analytical methods including suggestions for the range
of acceptable recovery rates.
AFLATOXIN TEST PROCEDURE AND MAXIMUM LEVELS
6. An aflatoxin sampling plan is defined by an aflatoxin test procedure and a maximum level. A value for
the ML and the aflatoxin test procedure are given below in this section.
7. The ML for “RTE” dried figs is 10 ng/g total aflatoxins.
8. Choice of the number and size of the laboratory sample is a compromise between minimizing risks (false
positives and false negatives) and costs related to sampling and restricting trade. For simplicity, it is
recommended that the aflatoxin sampling plan uses three 10 kg aggregate samples of dried figs.
9. The RTE sampling plan has been designed for enforcement and controls concerning total aflatoxins in
bulk consignments (lots) of dried figs traded in the export market.
Maximum level – 10 µg/kg total aflatoxins
Number of laboratory samples – 3
Laboratory sample size – 10 kg
Sample preparation – water-slurry grind and a test portion that represents 55 g mass of
dried figs
Analytical method – performance based (see Table 2)
Decision rule – If the aflatoxin test result is less than or equal to 10 µg/kg total
aflatoxins for all three 10 kg laboratory samples, then accept the lot.
Otherwise, reject the lot.
10. To assist Member Countries implement the above sampling plan, sample selection methods, sample
preparation methods, and analytical methods required to quantify aflatoxin in laboratory samples taken
from bulk dried fig lots are described in the following sections.
SAMPLE SELECTION
M ATERIAL TO BE SAMPLED
11. Each lot, which is to be examined for aflatoxin, must be sampled separately. Lots larger than 15 tonnes
should be subdivided into sublots to be sampled separately. If a lot is greater than 15 tonnes, the number
of sublots is equal to the lot weight in tonnes divided by 15 tonnes. It is recommended that a lot or a
sublot should not exceed 15 tonnes.
12. Considering that the weight of the lot is not always an exact multiple of 15 tonnes, the weight of the
sublot may exceed the mentioned weight by a maximum of 25 percent.
13. Samples should be taken from the same lot, i.e. they should have the same batch code or at the very
least the same best before date. Any changes, which would affect the mycotoxin content, the analytical
determination or make the aggregate samples collected unrepresentative should be avoided. For
example, do not open packaging in adverse weather conditions or expose samples to excessive
moisture or sunlight. Avoid cross-contamination from other potentially contaminated consignments
nearby.
14. In most cases any truck or container will have to be unloaded to allow representative sampling to be
carried out.
INCREMENTAL SAMPLE SELECTION
15. Procedures used to take incremental samples from a dried fig lot are extremely important. Every
individual fig in the lot should have an equal chance of being chosen. Biases will be introduced by
sample selection methods if equipment and procedures used to select the incremental samples prohibit
or reduce the chances of any item in the lot from being chosen.
16. Since there is no way to know if the contaminated figs are uniformly dispersed throughout the lot, it is
essential that the aggregate sample be the accumulation of many small incremental samples of product
selected from different locations throughout the lot. If the aggregate sample is larger than desired, it
should be blended and subdivided until the desired laboratory sample size is achieved.
17. For lots less than 10 tonnes, the size of the aggregate sample is reduced so that the aggregate sample
size does not exceed a significant portion of the lot or sublot size.
CXS 193-1995 30
NUMBER AND SIZE OF INCREMENTAL SAMPLES FOR LOTS OF VARYING WEIGHT
18. The number of incremental samples to be taken from a lot (sublot) depends on the weight of the lot.
Table 1 shall be used to determine the number of incremental samples to be taken from lots or sublots
of various sizes. The number of incremental samples varies from 10 to 100 for lots or sublots of various
sizes.
Table 1. Number and size of incremental samples composited
for an aggregate sample of 30 kga as a function of lot (or sublot) weight
Minimum Minimum Minimum Laboratory
Lot or sublot Number of
number of incremental aggregate sample
weightb laboratory
incremental sample sizec sample size size
(T in tonnes) samples
samples (g) (Kg) (Kg)
15.0 T > 10.0 100 300 30 10 3
10.0 T > 5.0 80 300 24 8 3
5.0 T > 2.0 60 300 18 9 2
2.0 T > 1.0 40 300 12 6 2
1.0 T > 0.5 30 300 9 9 1
0.5 T > 0.2 20 300 6 6 1
0.2 T > 0.1 15 300 4.5 4.5 1
0.1 T 10 300 3 3 1
Notes:
a Minimum aggregate sample size = laboratory sample size of 30 kg for lots above 10 tonnes
b1 Tonne = 1 000 kg
c Minimum incremental sample size = laboratory sample size (30 kg)/minimum number of
incremental samples, i.e. for 10 < T ≤ 15 tonnes,
300 g = 30 000/100
19. The suggested minimum weight of the incremental sample is 300 g for lots and sublots of various sizes.
STATIC LOTS
20. A static lot can be defined as a large mass of dried figs contained either in a large single container such
as a wagon, truck, or railcar or in many small containers such as sacks or boxes and the dried figs are
stationary at the time a sample is selected. Selecting a truly random sample from a static lot can be
difficult because all containers in the lot or sublot may not be accessible.
21. Taking incremental samples from a static lot usually requires the use of probing devices to select product
from the lot. The probing devices should be specifically designed for the commodity and type of
container. The probe should 1) be long enough to reach all products; 2) not restrict any item in the lot
from being selected; and 3) not alter the items in the lot. As mentioned above, the aggregate sample
should be a composite from many small incremental samples of product taken from many different
locations throughout the lot.
22. For lots traded in individual packages, the sampling frequency (SF), or number of packages that
incremental samples are taken from, is a function of the lot weight (LT), incremental sample weight (IS),
aggregate sample weight (AS) and the individual packing weight (IP), as follows:
Equation 1: SF = (LT x IS) / (AS x IP)
23. The SF is the number of packages sampled. All weights should be in the same mass units, such as kg.
DYNAMIC LOTS
24. Representative aggregate samples can be more easily produced when selecting incremental samples
from a moving stream of dried figs as the lot is transferred from one location to another. When sampling
from a moving stream, take small incremental samples of product from the entire length of the moving
stream; composite the incremental samples to obtain an aggregate sample; if the aggregate sample is
larger than the required laboratory sample(s), then blend and subdivide the aggregate sample to obtain
the desired size laboratory sample(s).
CXS 193-1995 31
25. Automatic sampling equipment such as a cross-cut sampler is commercially available with timers that
automatically pass a diverter cup through the moving stream at predetermined and uniform intervals.
When automatic sampling equipment is not available, a person can be assigned to manually pass a cup
through the stream at periodic intervals to collect incremental samples. Whether using automatic or
manual methods, incremental samples should be collected and composited at frequent and uniform
intervals throughout the entire time the figs flow past the sampling point.
26. Cross-cut samplers should be installed in the following manner: (1) the plane of the opening of the
diverter cup should be perpendicular to the direction of the flow; (2) the diverter cup should pass through
the entire cross-sectional area of the stream; and (3) the opening of the diverter cup should be wide
enough to accept all items of interest in the lot. As a general rule, the width of the diverter cup opening
should be about two to three times the largest dimensions of items in the lot.
27. The size of the aggregate sample (S) in kg, taken from a lot by a cross-cut sampler, is:
Equation 2: S = (D x LT) / (T x V)
where D is the width of the diverter cup opening (cm), LT is the lot size (kg), T is the interval or time
between cup movement through the stream (seconds), and V is cup velocity (cm/sec).
28. If the mass flow rate of the moving stream, MR (kg/sec), is known, then the sampling frequency (SF), or
number of cuts made by the automatic sampler cup can be computed from Equation 3 as a function of
S, V, D, and MR.
Equation 3: SF = (S x V) / (D x MR)
29. Equations 2 and 3 can also be used to compute other terms of interest such as the time between cuts
(T). For example, the time (T) required between cuts of the diverter cup to obtain a 30 kg aggregate
sample from a 20 000 kg lot where the diverter cup width is 5.0 cm and the cup velocity through the
stream 20 cm/sec. Solving for T in Equation 2.
T = (5.0 cm x 20 000 kg) / (30 kg x 20 cm/sec) = 167 sec
30. If the lot is moving at 500 kg per minute, the entire lot will pass through the sampler in 40 minutes
(2 400 sec) and only 14.4 cuts (14 incremental samples) will be made by the cup through the lot
(Equation 3). This may be considered too infrequent, in that too much product (1 388.9 kg) passes
through the sampler between the time the cup cuts through the stream.
PACKAGING AND TRANSPORTATION OF SAMPLES
31. Each laboratory sample shall be placed in a clean, inert container offering adequate protection from
contamination, sunlight and against damage in transit. All necessary precautions shall be taken to avoid
any change in composition of the laboratory sample, which might arise during transportation or storage.
Samples should be stored in a cool dark place.
SEALING AND LABELLING OF SAMPLES
32. Each laboratory sample taken for official use shall be sealed at the place of sampling and identified. A
record must be kept of each sampling, permitting each lot to be identified unambiguously and giving the
date and place of sampling together with any additional information likely to be of assistance to the
analyst.
SAMPLE PREPARATION
PRECAUTIONS
33. Sunlight should be excluded as much as possible during sample preparation, since aflatoxin gradually
breaks down under the influence of ultraviolet light. Also, environmental temperature and relative
humidity should be controlled and not favour mould growth and aflatoxin formation.
HOMOGENIZATION – GRINDING
34. As the distribution of aflatoxin is extremely non-homogeneous, the laboratory samples should be
homogenized by grinding the entire laboratory sample received by the laboratory. Homogenization is a
procedure that reduces particle size and disperses the contaminated particles evenly throughout the
comminuted laboratory sample.
35. The laboratory sample should be finely ground and mixed thoroughly using a process that approaches
as complete homogenization as possible. Complete homogenization implies that particle size is
extremely small, and the variability associated with sample preparation approaches zero. After grinding,
the grinder should be cleaned to prevent aflatoxin cross-contamination.
36. The use of vertical cutter mixer type grinders that mix and comminute the laboratory sample into a paste
represent a compromise in terms of cost and fineness of grind or particle size reduction. A better
homogenization (finer grind), such as a liquid slurry, can be obtained by more sophisticated equipment
and should provide the lowest sample preparation variance.
CXS 193-1995 32
TEST PORTION
37. The suggested weight of the test portion taken from the comminuted laboratory sample should be
approximately 50 g. If the laboratory sample is prepared using a liquid slurry, the slurry should contain
50 g of fig mass.
38. Procedures for selecting the 50 g test portion from the comminuted laboratory sample should be a
random process. If mixing occurred during or after the comminution process, the 50 g test portion can
be selected from any location throughout the comminuted laboratory sample. Otherwise, the 50 g test
portion should be the accumulation of several small portions selected throughout the laboratory sample.
39. It is suggested that three test portions be selected from each comminuted laboratory sample. The three
test portions will be used for enforcement, appeal and confirmation if needed.
ANALYTICAL METHODS
BACKGROUND
40. A criteria-based approach, whereby a set of performance criteria is established with which the analytical
method used should comply, is appropriate. The criteria-based approach has the advantage that, by
avoiding setting down specific details of the method used, developments in methodology can be
exploited without having to reconsider or modify the specific analytical method. The performance criteria
established for analytical methods should include all the parameters that need to be addressed by each
laboratory such as the detection limit, repeatability coefficient of variation (within lab), reproducibility
coefficient of variation (among laboratories), and the percent recovery necessary for various statutory
limits. Analytical methods that are accepted by chemists internationally such as the Association of
Official Analytical Collaboration (AOAC International) may be used. These methods are regularly
monitored and improved depending upon technology.
PERFORMANCE CRITERIA FOR METHODS OF ANALYSIS
41. A list of criteria and performance levels are shown in Table 2. Utilizing this approach, laboratories would
be free to use the analytical method most appropriate for their facilities.
Table 2. Specific requirements with which methods of analysis should comply with
Concentration Recommended
Criterion Maximum permitted value
range (ng/g) value
Blanks All Negligible n/a
1 to 15 70 to 100% n/a
Recovery
> 15 80 to 110% n/a
Precision or relative 1 to 120 Equation 4 2 x value derived from Equation 4
standard deviation
RSDR > 120 Equation 5 2 x value derived from Equation 5
(reproducibility)
Calculated as
1 to 120 0.66 times n/a
Precision or relative Precision RSDR
standard deviation
RSDr (repeatability) Calculated as
> 120 0.66 times n/a
Precision RSDr
Note: n/a = not applicable
42. The detection limits of the methods used are not stated. Only the precision values are given at the
concentrations of interest. The precision values (expressed as a%) are calculated from equations 4
and 5.
Equation 4: RSDR = 22.0
Equation 5: RSDR = 45.25C-0.15
where:
 RSDR = the relative standard deviation calculated from results generated under reproducibility
conditions
 RSDr = the relative standard deviation calculated from results generated under repeatability
conditions = 0.66RSDR
 C = aflatoxin concentration or mass of aflatoxin to mass of dried figs (i.e. ng/g)
CXS 193-1995 33
43. Equations 4 and 5 are generalized precision equations, which have been found to be independent of
analyte and matrix but solely dependent on concentration for most routine methods of analysis.
44. Results should be reported on the sample.
UNCERTAINTY, AS MEASURED BY THE VARIANCE, ASSOCIATED WITH THE SAMPLING,
SAMPLE PREPARATION, AND ANALYTICAL STEPS OF THE AFLATOXIN TEST PROCEDURE
USED TO DETECT AFLATOXIN IN DRIED FIGS
45. The sampling, sample preparation, and analytical variances associated with the aflatoxin test procedure
for dried figs are shown in Table 3.
Table 3. Variancesa associated with the aflatoxin test procedure for dried figs

Test Procedure Variances for Dried Figs

Samplingb,c S2s = (590/ns) 2.219C1.433

Sample prepd S2sp = (55/nss) 0.01170C1.465

S2a = (1/na) 0.0484C2.0
Analyticale

Total S2t = S2s + S2sp + S2a

Notes
aVariance = S2 (t, s, sp, and a denote total, sampling, sample preparation and analytical steps, respectively, of
aflatoxin test procedure)
b ns = laboratory sample size in number of dried figs, nss = test portion size in grams of fig mass, na = number of
aliquots quantified by HPLC, and C = aflatoxin concentration in ng/g total aflatoxins
c Count/kg for dried figs averaged 59/kg
d Sample preparation variance reflects a water-slurry method and a test portion that reflects 55 g fig mass
e Analytical variances reflect FAPAS recommendation for upper limit of analytical reproducibility uncertainty. A
relative standard deviation of 22 percent is based upon FAPAS data and considered as an appropriate measure of
the best agreement that can be obtained between laboratories. An analytical uncertainty of 22 percent is larger than
the within laboratory uncertainty measured in the sampling studies for the three dried figs.
CXS 193-1995 34
Annex VI
SAMPLING PLANS FOR TOTAL AFLATOXINS IN CERTAIN CEREALS AND
CEREAL-BASED PRODUCTS INCLUDING FOODS FOR INFANTS AND YOUNG CHILDREN
Sampling plans and performance criteria for aflatoxin (AFB1+AFB2+AFG1+AFG2) in maize grain,
destined for further processing

Maximum level 15 µg/kg AFB1+AFB2+AFG1+AFG2

Increments of 100 g, depending on the lot weight

Increments
(> 0.5 tonnes)

Dry grind with a suitable mill

Sample preparation
(particles smaller than 0.85 mm – 20 mesh)

Laboratory sample weight ≥ 5 kg

Number of laboratory
1
samples

Test portion 25 g

Selected according to the established performance criteria

Method
in Table 3.

If the sum of test results of AFB1, AFB2, AFG1 and AFG2

Decision rule for the laboratory sample is equal to or less than 15 µg/kg,
accept the lot. Otherwise, reject the lot.

Sampling plans and performance criteria for aflatoxin (AFB1+AFB2+AFG1+AFG2) in flour meal,
semolina and flakes derived from maize

Maximum level 10 µg/kg AFB1+AFB2+AFG1+AFG2

Increments 10 x 100g

Dry grind with a suitable mill

Sample preparation (particles smaller than 0.85 mm – 20 mesh),
if necessary for coarse samples

Laboratory sample weight 1 kg

Number of laboratory
1
samples

Test portion 25 g

Selected according to the established performance criteria

Method
in Table 3.

If the sum of test results of AFB1, AFB2, AFG1 and AFG2

Decision rule for the laboratory sample is equal to or less than 10 µg/kg,
accept the lot. Otherwise, reject the lot.
CXS 193-1995 35
Sampling plans and performance criteria for aflatoxin (AFB1+AFB2+AFG1+AFG2) in husked rice

Maximum level 20 µg/kg AFB1+AFB2+AFG1+AFG2

Increments of 100 g, depending on the lot weight

Increments
( >0.5 tonnes)

Dry grind with a suitable mill

Sample preparation
(particles smaller than 0.85 mm – 20 mesh)

Laboratory sample weight ≥ 5 kg

Number of laboratory
1
samples

Test portion 25 g

Selected according to the established performance criteria

Method
in Table 3.

If the sum of test results of AFB1, AFB2, AFG1, and AFG2

Decision rule for the laboratory sample is equal to or less than 20 µg/kg,
accept the lot. Otherwise, reject the lot.

Sampling plans and performance criteria for aflatoxin (AFB1+AFB2+AFG1+AFG2) in polished rice

Maximum level 5 µg/Kg AFB1+AFB2+AFG1+AFG2

Increments of 100 g, depending on the lot weight

Increments
(> 0.5 tonnes)

Dry grind with a suitable mill

Sample preparation
(particles smaller than 0.85 mm – 20 mesh)

Laboratory sample weight ≥ 5 kg

Number of laboratory
1
samples

Test portion 25 g

Selected according to the established performance criteria

Method
in Table 3.

If the sum of test results of AFB1, AFB2, AFG1 and AFG2

Decision rule for the laboratory sample is equal to or less than 5 µg/kg,
accept the lot. Otherwise, reject the lot.
CXS 193-1995 36
Sampling plans and performance criteria for aflatoxin (AFB1+AFB2+AFG1+AFG2) in sorghum

Maximum level 10 µg/kg AFB1+AFB2+AFG1+AFG2

Increments of 100 g, depending on the lot weight

Increments
(> 0.5 tonnes)

Dry grind with a suitable mill

Sample preparation
(particles smaller than 0.85 mm – 20 mesh)

Laboratory sample size ≥ 5 kg

Number of laboratory weight 1

Test portion 25 g

Selected according to the established performance criteria

Method
in Table 3.

If the sum of test results of AFB1, AFB2, AFG1 and AFG2

Decision rule for the laboratory sample is equal to or less than 10 µg/kg,
accept the lot. Otherwise, reject the lot.

Sampling plans and performance criteria for aflatoxin (AFB1+AFB2+AFG1+AFG2) in cereal-based food
for infants and young children

Maximum level 5 µg/kg AFB1+AFB2+AFG1+AFG2

Increments 10 x 100g

Dry grind with a suitable mill

Sample preparation (particles smaller than 0.85 mm – 20 mesh),
if necessary for coarse samples

Laboratory sample weight 1 kg

Number of laboratory
1
samples

Test portion 25 g

Selected according to the established performance criteria

Method
in Table 3.

If the sum of test results of AFB1, AFB2, AFG1 and AFG2

Decision rule for the laboratory sample is equal to or less than 5 µg/kg,
accept the lot. Otherwise, reject the lot.
CXS 193-1995 37
Sampling plans and performance criteria for aflatoxin (AFB1+AFB2+AFG1+AFG2) in cereal-based
food for infants and young children destined for food aid programmes

Maximum level 10 µg/kg AFB1+AFB2+AFG1+AFG2

Increments 10 x 100g

Dry grind with a suitable mill

Sample preparation (particles smaller than 0.85 mm – 20 mesh),
if necessary for coarse samples

Laboratory sample size 1 kg

Number of laboratory weight 1

Test portion 25 g

Selected according to the established performance criteria

Method
in Table 3.

If the sum of test results of AFB1, AFB2, AFG1 and AFG2

Decision rule for the laboratory sample is equal to or less than 10 µg/kg,
accept the lot. Otherwise, reject the lot.

DEFINITIONS

An identifiable quantity of a food commodity delivered at

one time and determined by the official to have common
Lot
characteristics, such as origin, variety, type of packing,
packer, consignor or markings.

Designated part of a larger lot in order to apply the

Sublot sampling method on that designated part. Each sublot must
be physically separate and identifiable.

It is defined by an aflatoxin test procedure and an

accept/reject level. An aflatoxin test procedure consists of
three steps: sample selection, sample preparation and
Sampling plan
analysis or aflatoxin quantification. The accept/reject level is
a tolerance usually equal to the Codex maximum level
(ML).

The quantity of material taken from a single random place

Incremental sample
in the lot or sublot.

The combined total of all the incremental samples that is

taken from the lot or sublot. The aggregate sample has to
Aggregate sample
be at least as large as the laboratory sample or samples
combined.

The smallest quantity of cereal grains, shelled cereal

grains, and cereal-based products comminuted in a mill.
The laboratory sample may be a portion of or the entire
aggregate sample. If the aggregate sample is larger than
Laboratory sample
the laboratory sample/s, the laboratory sample/s should be
removed in a random manner from the aggregate sample in
such a way to ensure that the laboratory sample is still
representative of the sublot sampled.

A portion of the comminuted laboratory sample. The entire

laboratory sample should be comminuted in a mill. A portion
Test portion
of the comminuted laboratory sample is randomly removed
for the extraction of the aflatoxin for chemical analysis.
CXS 193-1995 38
SAMPLING PLAN DESIGN CONSIDERATIONS
M ATERIAL TO BE SAMPLED
1. Each lot of cereal grains and cereal-based products which is to be examined for AFs must be sampled
separately. Lots larger than 50 tonnes should be subdivided into sublots to be sampled separately. If a
lot is greater than 50 tonnes, the lot should be subdivided into sublots according to Table 1.
Table 1. Subdivision of cereal grains sublots according to lot weight – maize grain, sorghum, polished
rice and husked rice

Maximum weight or Number of Minimum

Lot weight (t) minimum number incremental laboratory sample
of sublots samples weight (kg)

> 1 500 500 tonnes 100 5

> 300 and < 1 500 3 sublots 100 5

> 100 and ≤ 300 100 tonnes 100 5

> 50 and < 100 2 sublots 100 5

< 50 - 3–100* 5

Note: *See Table 2

2. Considering that the weight of the lot is not always an exact multiple of the weight of sublots, the weight
of the sublot may exceed the mentioned size by a maximum of 20 percent.
INCREMENTAL SAMPLE
3. The suggested minimum size of the incremental sample of cereal grains and cereal-based products
should be 100 g for lots ≥ 0.5 tonnes.
4. For lots less than 50 tonnes of cereal grains and cereal-based products, the sampling plan must be used
with 3 to 100 incremental samples, depending on the lot weight. For very small lots (< 0.5 tonnes), a
lower number of incremental samples may be taken, but the aggregate sample uniting all incremental
samples shall also be, in that case, at least 5 kg. Table 2 may be used to determine the number of
incremental samples to be taken.
Table 2. Number of incremental samples of cereal grains to be taken depending on the weight of the
lot- Maize grain, sorghum, polished rice and husked rice

Number of incremental Minimum laboratory

Lot weight (t)
samples sample weight (kg)

<0.05 3 5

>0.05–< 0.5 5 5

>0.5–< 1 10 5

>1–< 3 20 5

>3–< 10 40 5

>10–< 20 60 5

>20–< 50 100 5
CXS 193-1995 39
STATIC LOTS
5. A static lot can be defined as a large mass of cereal grains and cereal-based products contained either
in a large single container such as a wagon, truck, or railcar or in many small containers such as sacks
or boxes, and the cereal grains and cereal-based products is stationary at the time a sample is selected.
Selecting a truly random sample from a static lot can be difficult because all containers in the lot or
sublot may not be accessible.
6. Taking incremental samples from a static lot usually requires the use of probing devices to select product
from the lot. The probing devices should be specifically designed for the commodity and type of
container. The probe should: 1) be long enough to reach all products; 2) not restrict any item in the lot
from being selected; and 3) not alter the items in the lot. As mentioned above, the aggregate sample
should be a composite from many small incremental samples of product taken from many different
locations throughout the lot.
7. For lots traded in individual packages, the sampling frequency (SF), or number of packages that
incremental samples are taken from, is a function of the lot size (LT), incremental sample size (IS),
aggregate sample size (AS) and the individual packing size (IP), as follows:
S = (LT x IS) / (AS x IP)
8. The SF is the number of packages sampled. All sizes should be in the same mass units, such as kg.
DYNAMIC LOTS
9. Representative aggregate samples can be more easily produced when selecting incremental samples
from a moving stream of cereal grains and cereal-based products as the lot is transferred from one
location to another. When sampling from a moving stream, take small incremental samples of product
from the entire length of the moving stream; composite the incremental samples to obtain an aggregate
sample; if the aggregate sample is larger than the required laboratory sample/s, then blend and
subdivide the aggregate sample to obtain the desired size laboratory sample/s.
10. Automatic sampling equipment such as a cross-cut sampler is commercially available with timers that
automatically pass a diverter cup through the moving stream at predetermined and uniform intervals.
When automatic sampling equipment is not available, a person can be assigned to manually pass a cup
through the stream at periodic intervals to collect incremental samples. Whether using automatic or
manual methods, incremental samples should be collected and composited at frequent and uniform
intervals throughout the entire time the cereal flow past the sampling point.
11. Cross-cut samplers should be installed in the following manner: 1) the plane of the opening of the
diverter cup should be perpendicular to the direction of the flow; 2) the diverter cup should pass through
the entire cross-sectional area of the stream; and 3) the opening of the diverter cup should be wide
enough to accept all items of interest in the lot. As a general rule, the width of the diverter cup opening
should be about two to three times the largest dimensions of items in the lot.
12. The size of the aggregate sample (S) in kg, taken from a lot by a cross-cut sampler is:
S = (D x LT) / (T x V),
where D is the width of the diverter cup opening (cm), LT is the lot size (kg), T is interval or time between
cup movement through the stream (seconds), and V is cup velocity (cm/sec).
13. If the mass flow rate of the moving stream, MR (kg/sec), is known, then the SF, or number of cuts made
by the automatic sampler cup can be computed as a function of S, V, D, and MR.
SF = (S x V) / (D x MR).
PACKAGING AND TRANSPORTATION OF SAMPLES
14. Each laboratory sample shall be placed in a clean, inert container offering adequate protection from
contamination, sunlight and against damage in transit. All necessary precautions shall be taken to avoid
any change in composition of the laboratory sample, which might arise during transportation or storage.
Samples should be stored in a cool dark place.
SEALING AND LABELLING OF SAMPLES
15. Each laboratory sample taken for official use shall be sealed at the place of sampling and identified. A
record must be kept of each sampling, permitting each lot to be identified unambiguously and giving the
date and place of sampling together with any additional information likely to be of assistance to the
analyst.
CXS 193-1995 40
SAMPLE PREPARATION PRECAUTIONS
16. Sunlight should be excluded as much as possible during sample preparation since aflatoxin gradually
breaks down under the influence of ultraviolet light. Also, environmental temperature and relative
humidity should be controlled and not favour mould growth and aflatoxin formation.
HOMOGENIZATION – GRINDING
17. As the distribution of aflatoxin is extremely non-homogeneous, laboratory samples should be
homogenized by grinding the entire laboratory sample received by the laboratory. Homogenization is a
procedure that reduces particle size and disperses the contaminated particles evenly throughout the
comminuted laboratory sample.
18. The laboratory sample should be finely ground and mixed thoroughly using a process that approaches
as complete homogenization as possible. Complete homogenization implies that particle size is
extremely small, and the variability associated with sample preparation is minimized. After grinding, the
grinder should be cleaned to prevent aflatoxin cross-contamination.
TEST PORTION
19. The suggested weight of the test portion taken from the comminuted laboratory sample should be
approximately 25 g. If the laboratory sample is prepared using a liquid slurry, the slurry should contain
25 g.
20. Procedures for selecting the 25 g test portion from the comminuted laboratory sample should be a
random process. If mixing occurred during or after the comminution process, the 25 g test portion can
be selected from any location throughout the comminuted laboratory sample. Otherwise, the 25 g test
portion should be the accumulation of several small portions selected throughout the laboratory sample.
ANALYTICAL METHODS
21. A criteria-based approach, whereby a set of performance criteria is established with which the analytical
method used should comply, is appropriate. The criteria-based approach has the advantage that, by
avoiding setting down specific details of the method used, developments in methodology can be
exploited without having to reconsider or modify the specific method. A list of possible criteria and
performance levels is shown in Table 3. Utilizing this approach, laboratories would be free to use the
analytical method most appropriate for their facilities.
 CXS 193-1995 41
Table 3. Method criteria for total aflatoxins in cereals, considering AFB1:AFB2:AFG1:AFG2 of 1:1:1:1.

Minimal
ML LOD LOQ Precision Recovery
Commodity Analyte applicable
(µg/kg) (µg/kg) (µg/kg) (%) (%)
range (µg/kg)
Maize grain AF B1+B2+G1+G2 15 ≤3 ≤6 < 44 8.4 – 21.6 60 – 115
AFB1 - ≤ 0.75 ≤ 1.5 < 44 2.1 – 5.4 40 – 120
AFB2 - ≤ 0.75 ≤ 1.5 < 44 2.1 – 5.4 40 – 120
AFG1 - ≤ 0.75 ≤ 1.5 < 44 2.1 – 5.4 40 – 120
AFG2 - ≤ 0.75 ≤ 1.5 < 44 2.1 – 5.4 40 – 120

Maize flour,
meal, semolina
and flakes
derived from
maize;
sorghum grain;
AF B1+B2+G1+G2 10 ≤2 ≤4 < 44 5.6 – 14.4 60 – 115
cereal-based
foods for
infants and
young children
for food aid
programmes
AFB1 - ≤ 0.5 ≤ 1.0 < 44 1.4 – 3.6 40 – 120
AFB2 - ≤ 0.5 ≤ 1.0 < 44 1.4 – 3.6 40 – 120
AFG1 - ≤ 0.5 ≤ 1.0 < 44 1.4 – 3.6 40 – 120
AFG2 - ≤ 0.5 ≤ 1.0 < 44 1.4 – 3.6 40 – 120

Husked rice AF B1+B2+G1+G2 20 ≤4 ≤8 < 44 11.2 – 28.8 60 – 115

AFB1 - ≤ 1.0 ≤ 2.0 < 44 2.8 – 7.2 40 – 120
AFB2 - ≤ 1.0 ≤ 2.0 < 44 2.8 – 7.2 40 – 120
AFG1 - ≤ 1.0 ≤ 2.0 < 44 2.8 – 7.2 40 – 120
AFG2 - ≤ 1.0 ≤ 2.0 < 44 2.8 – 7.2 40 – 120

Polished rice;
cereal-based
food for infants AF B1+B2+G1+G2 5 ≤1 ≤2 < 44 2.8 – 7.2 40 – 120
and young
children
AFB1 - ≤ 0.25 ≤ 0.5 < 44 0.7 – 1.8 40 – 120
AFB2 - ≤ 0.25 ≤ 0.5 < 44 0.7 – 1.8 40 – 120
AFG1 - ≤ 0.25 ≤ 0.5 < 44 0.7 – 1.8 40 – 120
AFG2 - ≤ 0.25 ≤ 0.5 < 44 0.7 – 1.8 40 – 120
CXS 193-1995 42
AFLATOXIN M1
Reference to JECFA: 56 (2001)
Toxicological guidance value: Cancer potency estimates at specified residue levels (2001, using worst-case assumptions, the additional risks for
liver cancer predicted with use of proposed maximum levels of aflatoxin M 1 of 0.05 and 0.5 µg/kg are very small.
The potency of aflatoxin M1 appears to be so low in HBsAg- individuals that a carcinogenic effect of M1 intake in
those who consume large quantities of milk and milk products in comparison with non-consumers of these products
would be impossible to demonstrate. Hepatitis B virus carriers might benefit from a reduction in the aflatoxin
concentration in their diet, and the reduction might also offer some protection in hepatitis C virus carriers).
Contaminant definition: Aflatoxin M1
Synonyms: AFM1
Related code of practice: Code of practice for the reduction of aflatoxin B1 in raw materials and supplemental feeding stuffs for milk producing
animals (CXC 45-1997)8
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name µg/kg product to which the ML applies
Milk is the normal mammary secretion of milking animals obtained from
one or more milkings without either addition to it or extraction from it,
Milks 0.5 Whole commodity intended for consumption as liquid milk or for further processing.
A concentration factor applies to partially or wholly dehydrated milks.
CXS 193-1995 43
DEOXYNIVALENOL (DON)
Reference to JECFA: 56 (2001), 72 (2010)
Toxicological guidance value: Group PMTDI 0.001 mg/kg bw (2010, for DON and its acetylated derivates)
Group ARfD 0.008 mg/kg bw (2010, for DON and its acetylated derivates)
Contaminant definition: Deoxynivalenol
Synonyms: Vomitoxin; abbreviation, DON
Related code of practice: Code of practice for the prevention and reduction of mycotoxin contamination in cereals (CXC 51-2003)11
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name µg/kg product to which the ML applies
The relevant Codex commodity standard is CXS 74-1981.
Cereal-based foods Whole commodity as sold; not
The ML applies to all cereal-based foods intended for infants
for infants and young 200 reconstituted or otherwise prepared
(up to 12 months) and young children (12 to 36 months)
children for consumption
For sampling plan, see Annex VII.
Flour, meal, semolina
and flakes derived
1 000 For sampling plan, see Annex VII.
from wheat, maize or
barley
“Destined for further processing” means intended to undergo an
Cereal grains (wheat, additional processing/treatment that has proven to reduce levels of DON
maize and barley) before being used as an ingredient in foodstuffs otherwise processed or
2 000 offered for human consumption. Codex members may define the
destined for further
processing processes that have been shown to reduce levels.
For sampling plan, see Annex VII.
CXS 193-1995 44

Annex VII
SAMPLING PLANS AND PERFORMANCE CRITERIA FOR DEOXYNIVALENOL (DON) IN
CEREAL-BASED FOODS FOR INFANTS AND YOUNG CHILDREN;
IN FLOUR, MEAL, SEMOLINA AND FLAKES DERIVED FROM WHEAT, MAIZE OR BARLEY; AND IN
CEREAL GRAINS (WHEAT, MAIZE AND BARLEY) DESTINED FOR FURTHER PROCESSING
Cereal grains (wheat, maize, and barley) destined for further processing
Maximum level 2 000 µg/kg DON
Increments increments of 100 g, depending on the lot weight (≥ 0.5 tonnes)
Sample preparation dry grind with a suitable mill (particles smaller than 0.85 mm–20 mesh)
Laboratory sample weight ≥ 1 kg
Number of laboratory samples 1
Test portion 25 g test portion
Method HPLC
Decision rule If the DON sample test result for the laboratory samples is equal to or
less than 2 000 µg/kg, accept the lot. Otherwise, reject the lot.
Cereal-based foods for infants and young children
Maximum level 200 µg/kg DON
Increments 10 x 100 g
Sample preparation None
Laboratory sample weight 1 kg
Number of laboratory samples 1
Test portion 25 g test portion
Method HPLC
Decision rule If the DON sample test result is equal or less than 200 µg/kg,
accept the lot. Otherwise, reject the lot.
Flour, semolina, meal and flakes derived from wheat, maize or barley
Maximum level 1 000 µg/kg DON
Increments 10 x 100 g
Sample preparation None
Laboratory sample weight 1 kg
Number of laboratory samples 1
Test portion 25 g test portion
Method HPLC
Decision rule If the DON sample test result is equal or less than
1 000 µg/kg, accept the lot. Otherwise, reject the lot.
DEFINITIONS
An identifiable quantity of a food commodity delivered at one time and determined
Lot by the official to have common characteristics, such as origin, variety, type of
packing, packer, consignor or markings.
Designated part of a larger lot in order to apply the sampling method on that
Sublot
designated part. Each sublot must be physically separate and identifiable.
It is defined by a DON test procedure and an accept/reject level. A DON test
procedure consists of three steps: sample selection, sample preparation and
Sampling plan
analysis or DON quantification. The accept/reject level is a tolerance usually equal
to the Codex maximum level (ML).
CXS 193-1995 45

Incremental sample The quantity of material taken from a single random place in the lot or sublot.
The combined total of all the incremental samples that is taken from the lot or
Aggregate sample sublot. The aggregate sample has to be at least as large as the laboratory sample
or samples combined.
The smallest quantity of shelled cereal comminuted in a mill. The laboratory
sample may be a portion of or the entire aggregate sample. If the aggregate
Laboratory sample sample is larger than the laboratory sample(s), the laboratory sample(s) should be
removed in a random manner from the aggregate sample in such a way to ensure
that the laboratory sample is still representative of the sublot sampled.
A portion of the comminuted laboratory sample. The entire laboratory sample
Test portion should be comminuted in a mill. A portion of the comminuted laboratory sample is
randomly removed for the extraction of the DON for chemical analysis.
SAMPLING PLAN DESIGN CONSIDERATIONS
M ATERIAL TO BE SAMPLED
1. Each lot of cereal, which is to be examined for DON, must be sampled separately. Lots larger than
50 tonnes should be subdivided into sublots to be sampled separately. If a lot is greater than 50 tonnes,
the lot should be subdivided into sublots according to Table 1.
Table 1. Subdivision of cereal sublots according to lot weight
Lot weight Maximum weight or minimum Number of Minimum laboratory
(t) number of sublots incremental samples sample weight (kg)
≥ 1 500 500 tonnes 100 1
> 300 and
3 sublots 100 1
< 1 500
≥ 100 and
100 tonnes 100 1
≤ 300
≥ 50 and
2 sublots 100 1
< 100
< 50 - 3 – 100* 1
Note: * See Table 2
2. Considering that the weight of the lot is not always an exact multiple of the weight of sublots, the weight
of the sublot may exceed the mentioned weight by a maximum of 20 percent.
INCREMENTAL SAMPLE
3. The suggested minimum weight of the incremental sample should be 100 grams for lots ≥ 0.5 tonnes.
4. For lots less than 50 tonnes, the sampling plan must be used with 3 to 100 incremental samples,
depending on the lot weight. For very small lots ( 0.5 tonnes, a lower number of incremental samples
may be taken, but the aggregate sample uniting all incremental samples shall also be, in that case, at
least 1 kg. Table 2 may be used to determine the number of incremental samples to be taken.
Table 2. Number of incremental samples to be taken depending on the weight of the lot
Minimum laboratory sample
Lot weight (t) Number of incremental samples
weight (kg)
≤ 0.05 3 1
> 0.05 – ≤ 0.5 5 1
> 0.5 – ≤ 1 10 1
>1–≤3 20 1
> 3 – ≤ 10 40 1
> 10 – ≤ 20 60 1
> 20 – < 50 100 1
CXS 193-1995 46

STATIC LOTS
5. A static lot can be defined as a large mass of shelled cereal contained either in a large single container
such as a wagon, truck or railcar or in many small containers such as sacks or boxes and the cereal is
stationary at the time a sample is selected. Selecting a truly random sample from a static lot can be
difficult because all containers in the lot or sublot may not be accessible.
6. Taking incremental samples from a static lot usually requires the use of probing devices to select product
from the lot. The probing devices should be specifically designed for the commodity and type of
container. The probe should: 1) be long enough to reach all products; 2) not restrict any item in the lot
from being selected; and 3) not alter the items in the lot. As mentioned above, the aggregate sample
should be a composite from many small incremental samples of product taken from many different
locations throughout the lot.
7. For lots traded in individual packages, the sampling frequency (SF), or number of packages that
incremental samples are taken from, is a function of the lot weight (LT), incremental sample weight (IS),
aggregate sample weight (AS) and the individual packing weight (IP), as follows:
SF = (LT x IS) / (AS x IP)
8. The SF is the number of packages sampled. All weights should be in the same mass units, such as kg.
DYNAMIC LOTS
9. Representative aggregate samples can be more easily produced when selecting incremental samples
from a moving stream of shelled cereal as the lot is transferred from one location to another. When
sampling from a moving stream, take small incremental samples of product from the entire length of the
moving stream; composite the incremental samples to obtain an aggregate sample; if the aggregate
sample is larger than the required laboratory sample(s), then blend and subdivide the aggregate sample
to obtain the desired size laboratory sample(s).
10. Automatic sampling equipment such as a cross-cut sampler is commercially available with timers that
automatically pass a diverter cup through the moving stream at predetermined and uniform intervals.
When automatic sampling equipment is not available, a person can be assigned to manually pass a cup
through the stream at periodic intervals to collect incremental samples. Whether using automatic or
manual methods, incremental samples should be collected and composited at frequent and uniform
intervals throughout the entire time the cereal flow past the sampling point.
11. Cross-cut samplers should be installed in the following manner: 1) the plane of the opening of the
diverter cup should be perpendicular to the direction of the flow; 2) the diverter cup should pass through
the entire cross-sectional area of the stream; and 3) the opening of the diverter cup should be wide
enough to accept all items of interest in the lot. As a general rule, the width of the diverter cup opening
should be about two to three times the largest dimensions of items in the lot.
12. The size of the aggregate sample (S) in kg, taken from a lot by a cross-cut sampler is:
S = (D x LT) / (T x V)
where D is the width of the diverter cup opening (cm), LT is the lot size (kg), T is interval or time between
cup movement through the stream (seconds), and V is cup velocity (cm/sec).
13. If the mass flow rate of the moving stream, MR (kg/sec), is known, then the SF, or number of cuts made
by the automatic sampler cup can be computed as a function of S, V, D, and MR.
SF = (S x V) / (D x MR)
PACKAGING AND TRANSPORTATION OF SAMPLES
14. Each laboratory sample shall be placed in a clean, inert container offering adequate protection from
contamination, sunlight and against damage in transit. All necessary precautions shall be taken to avoid
any change in composition of the laboratory sample, which might arise during transportation or storage.
Samples should be stored in a cool dark place.
15. Each laboratory sample taken for official use shall be sealed at the place of sampling and identified. A
record must be kept of each sampling, permitting each lot to be identified unambiguously and giving the
date and place of sampling together with any additional information likely to be of assistance to the
analyst.
SAMPLE PREPARATION
16. Sunlight should be excluded as much as possible during sample preparation, since DON may gradually
break down under the influence of ultraviolet light. Also, environmental temperature and relative humidity
should be controlled and not favour mould growth and DON formation.
CXS 193-1995 47

17. As the distribution of DON is extremely non-homogeneous, laboratory samples should be homogenized
by grinding the entire laboratory sample received by the laboratory. Homogenization is a procedure that
reduces particle size and disperses the contaminated particles evenly throughout the comminuted
laboratory sample.
18. The laboratory sample should be finely ground and mixed thoroughly using a process that approaches
as complete homogenization as possible. Complete homogenization implies that particle size is
extremely small, and the variability associated with sample preparation approaches zero. After grinding,
the grinder should be cleaned to prevent DON cross-contamination.
TEST PORTION
19. The suggested weight of the test portion taken from the comminuted laboratory sample should be
approximately 25 g
20. Procedures for selecting the test portion from the comminuted laboratory sample should be a random
process. If mixing occurred during or after the comminuting process, the test portion can be selected
from any location throughout the comminuted laboratory sample. Otherwise, the test portion should be
the accumulation of several small portions selected throughout the laboratory sample.
21. It is suggested that three test portions be selected from each comminuted laboratory sample. The three
test portions will be used for enforcement, appeal and confirmation if needed.
ANALYTICAL METHODS
22. A criteria-based approach, whereby a set of performance criteria is established with which the analytical
method used should comply, is appropriate. The criteria-based approach has the advantage that, by
avoiding setting down specific details of the method used, developments in methodology can be
exploited without having to reconsider or modify the specific method. A list of possible criteria and
performance levels are shown in Table 3). Utilizing this approach, laboratories would be free to use the
analytical method most appropriate for their facilities.
Table 3. Proposed method criteria for DON in cereals.

Minimum
ML LOD LOQ Precision on
Commodity applicable Recovery
(mg/kg) (mg/kg) (mg/kg) HorRat
range (mg/kg)
Cereal grains (wheat,
maize and barley)
2.0 ≤ 0.2 ≤ 0.4 ≤2 1–3 80 -–110%
destined for further
processing
Cereal-based foods
for infants and young 0.2 ≤ 0.02 ≤ 0.04 ≤2 0.1 – 0.3 80 – 110%
children
Flour, semolina,
meal and flakes
1.0 ≤ 0.1 ≤ 0.2 ≤2 0.5 – 1.5 80 – 110%
derived from wheat,
maize or barley
CXS 193-1995 48

FUMONISINS (B1 + B2)

Reference to JECFA: 56 (2001), 74 (2011)
Toxicological guidance value: PMTDI 0.002 mg/kg bw (2001, 2011)
Contaminant definition: Fumonisins (B1+ B2)
Synonyms: Several related compounds have been described, notably fumonisin B1, B2, and B3 (abbreviation: FB1, etc.)
Related code of practice: Code of practice for the prevention and reduction of mycotoxin contamination in cereals (CXC 51-2003)11
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name µg/kg product to which the ML applies

Raw maize grain 4 000 Whole commodity For sampling plan, see Annex VIII.

Maize flour and

2 000 Whole commodity For sampling plan, see Annex VIII.
maize meal
CXS 193-1995 49

Annex VIII
SAMPLING PLANS AND PERFORMANCE CRITERIA FOR FUMONISINS (FB1 + FB2)
IN MAIZE GRAIN AND MAIZE FLOUR AND MAIZE MEAL
Maize grain, unprocessed
Maximum level 4 000 µg/kg FB1 + FB2
Increments increments of 100 g, depending on the lot weight (≥ 0.5 tonnes)
dry grind with a suitable mill
Sample preparation
(particles smaller than 0.85 mm – 20 mesh)
Laboratory sample weight ≥ 1 kg
Number of laboratory samples 1
Test portion 25 g test portion
Method HPLC
If the fumonisin-sample test result for the laboratory samples is equal
Decision rule
to or less than 4 000 µg/kg, accept the lot. Otherwise, reject the lot.
Maize flour and maize meal
Maximum level 2 000 µg/kg FB1 + FB2
Increments 10 x 100 g
Sample preparation None
Laboratory sample weight ≥ 1 kg
Number of laboratory samples 1
Test portion 25 g test portion
Method HPLC
If the fumonisin-sample test result is equal or less than
Decision rule
2 000 µg/kg, accept the lot. Otherwise, reject the lot.
DEFINITION
An identifiable quantity of a food commodity delivered at one time and
Lot determined by the official to have common characteristics, such as origin,
variety, type of packing, packer, consignor or markings.
The designated part of a larger lot in order to apply the sampling method on
Sublot
that designated part. Each sublot must be physically separate and identifiable.
It is defined by a fumonisin test procedure and an accept/reject level. A
fumonisin test procedure consists of three steps: sample selection, sample
Sampling plan
preparation and analysis or fumonisin quantification. The accept/reject level is
a tolerance usually equal to the Codex maximum level (ML).
Incremental sample The quantity of material taken from a single random place in the lot or sublot.
The combined total of all the incremental samples that is taken from the lot or
Aggregate sample sublot. The aggregate sample has to be at least as large as the laboratory
sample or samples combined.
The smallest quantity of shelled maize comminuted in a mill. The laboratory
sample may be a portion of or the entire aggregate sample. If the aggregate
Laboratory sample sample is larger than the laboratory sample(s), the laboratory sample(s) should
be removed in a random manner from the aggregate sample in such a way to
ensure that the laboratory sample is still representative of the sublot sampled.
A portion of the comminuted laboratory sample. The entire laboratory sample
Test portion should be comminuted in a mill. A portion of the comminuted laboratory sample
is randomly removed for the extraction of the fumonisin for chemical analysis.
CXS 193-1995 50

SAMPLING PLAN DESIGN CONSIDERATIONS

M ATERIAL TO BE SAMPLED
1. Each lot of maize, which is to be examined for fumonisin, must be sampled separately. Lots larger than
50 tonnes should be subdivided into sublots to be sampled separately. If a lot is greater than 50 tonnes,
the lot should be subdivided into sublots according to Table 1.
Table 1. Subdivision of maize sublots according to lot weight
Maximum weight or
Number of Minimum laboratory
Lot weight (t) minimum number of
incremental sample sample weight (kg)
sublots
≥ 1 500 500 tonnes 100 1
> 300 and < 1 500 3 sublots 100 1
≥ 100 and ≤ 300 100 tonnes 100 1
≥ 50 and < 100 2 sublots 100 1
< 50 - 3–100* 1

Note: * See Table 2

2. Considering that the weight of the lot is not always an exact multiple of the weight of sublots, the weight
of the sublot may exceed the mentioned weight by a maximum of 20 percent.
INCREMENTAL SAMPLE
3. The suggested minimum weight of the incremental sample should be 100 grams for lots ≥ 0.5 tonnes.
4. For lots less than 50 tonnes, the sampling plan must be used with 3 to 100 incremental samples,
depending on the lot weight. For very small lots ( 0.5 tonnes, a lower number of incremental samples
may be taken, but the aggregate sample uniting all incremental samples shall also be, in that case at
least 1 kg. Table 2 may be used to determine the number of incremental samples to be taken.
Table 2. Number of incremental samples to be taken depending on the weight of the lot

Number of Minimum laboratory

Lot weight (t)
incremental sample sample weight (kg)
≤ 0.05 3 1
> 0.05 – ≤ 0.5 5 1
> 0.5 – ≤ 1 10 1
>1–≤3 20 1
> 3 – ≤ 10 40 1
> 10 – ≤ 20 60 1
> 20 – < 50 100 1
STATIC LOTS
5. A static lot can be defined as a large mass of shelled maize contained either in a large single container
such as a wagon, truck, or railcar or in many small containers such as sacks or boxes and the maize is
stationary at the time a sample is selected. Selecting a truly random sample from a static lot can be
difficult because all containers in the lot or sublot may not be accessible.
6. Taking incremental samples from a static lot usually requires the use of probing devices to select product
from the lot. The probing devices should be specifically designed for the commodity and type of
container. The probe should (1) be long enough to reach all products, (2) not restrict any item in the lot
from being selected, and (3) not alter the items in the lot. As mentioned above, the aggregate sample
should be a composite from many small incremental samples of product taken from many different
locations throughout the lot.
7. For lots traded in individual packages, the SF or number of packages that incremental samples are
taken from, is a function of the lot weight (LT), incremental sample weight (IS), aggregate sample weight
(AS) and the individual packing weight (IP), as follows:
SF = (LT x IS) / (AS x IP)
CXS 193-1995 51

8. The SF is the number of packages sampled. All weights should be in the same mass units, such as kg.
DYNAMIC LOTS
9. Representative aggregate samples can be more easily produced when selecting incremental samples
from a moving stream of shelled maize as the lot is transferred from one location to another. When
sampling from a moving stream, take small incremental samples of product from the entire length of the
moving stream; composite the incremental samples to obtain an aggregate sample; if the aggregate
sample is larger than the required laboratory sample(s), then blend and subdivide the aggregate sample
to obtain the desired size laboratory sample(s).
10. Automatic sampling equipment such as a cross-cut sampler is commercially available with timers that
automatically pass a diverter cup through the moving stream at predetermined and uniform intervals.
When automatic sampling equipment is not available, a person can be assigned to manually pass a cup
through the stream at periodic intervals to collect incremental samples. Whether using automatic or
manual methods, incremental samples should be collected and composited at frequent and uniform
intervals throughout the entire time the maize flow past the sampling point.
11. Cross-cut samplers should be installed in the following manner: 1) the plane of the opening of the
diverter cup should be perpendicular to the direction of the flow; 2) the diverter cup should pass through
the entire cross-sectional area of the stream; and 3) the opening of the diverter cup should be wide
enough to accept all items of interest in the lot. As a general rule, the width of the diverter cup opening
should be about two to three times the largest dimensions of items in the lot.
12. The size of the aggregate sample (S) in kg, taken from a lot by a cross-cut sampler is:
S = (D x LT) / (T x V)
where D is the width of the diverter cup opening (cm), LT is the lot size (kg), T is interval or time
between cup movement through the stream (seconds), and V is cup velocity (cm/sec).
13. If the mass flow rate of the moving stream, MR (kg/sec), is known, then the sampling frequency (SF), or
number of cuts made by the automatic sampler cup can be computed as a function of S, V, D, and MR.
SF = (S x V) / (D x MR)
PACKAGING AND TRANSPORTATION OF SAMPLES
14. Each laboratory sample shall be placed in a clean, inert container offering adequate protection from
contamination, sunlight and against damage in transit. All necessary precautions shall be taken to avoid
any change in composition of the laboratory sample, which might arise during transportation or storage.
Samples should be stored in a cool dark place.
15. Each laboratory sample taken for official use shall be sealed at the place of sampling and identified. A
record must be kept of each sampling, permitting each lot to be identified unambiguously and giving the
date and place of sampling together with any additional information likely to be of assistance to the
analyst.
SAMPLE PREPARATION
16. Sunlight should be excluded as much as possible during sample preparation, since fumonisin may
gradually break down under the influence of ultraviolet light. Also, environmental temperature and
relative humidity should be controlled and not favour mould growth and fumonisin formation.
17. As the distribution of fumonisin is extremely non-homogeneous, laboratory samples should be
homogenized by grinding the entire laboratory sample received by the laboratory. Homogenization is a
procedure that reduces particle size and disperses the contaminated particles evenly throughout the
comminuted laboratory sample.
18. The laboratory sample should be finely ground and mixed thoroughly using a process that approaches
as complete homogenization as possible. Complete homogenization implies that particle size is
extremely small, and the variability associated with sample preparation approaches zero. After grinding,
the grinder should be cleaned to prevent fumonisin cross-contamination.
TEST PORTION
19. The suggested weight of the test portion taken from the comminuted laboratory sample should be
approximately 25 g
20. Procedures for selecting the test portion from the comminuted laboratory sample should be a random
process. If mixing occurred during or after the comminuting process, the test portion can be selected
from any location throughout the comminuted laboratory sample. Otherwise, the test portion should be
the accumulation of several small portions selected throughout the laboratory sample.
CXS 193-1995 52

21. It is suggested that three test portions be selected from each comminuted laboratory sample. The three
test portions will be used for enforcement, appeal and confirmation if needed.
ANALYTICAL METHODS
22. A criteria-based approach, whereby a set of performance criteria is established with which the analytical
method used should comply, is appropriate. The criteria-based approach has the advantage that, by
avoiding setting down specific details of the method used, developments in methodology can be
exploited without having to reconsider or modify the specific method. A list of possible criteria and
performance levels are shown in Table 3. Utilizing this approach, laboratories would be free to use the
analytical method most appropriate for their facilities.
Table 3. Performance criteria for Fumonisin B1+ B2.
Maize Grain

Analyte ML (mg/Kg) LOD (mg/Kg) LOQ (mg/Kg) RSDR Recovery (%)

FB1 + FB2 4.0 - - - -
HorRat ≤ 2
FB1 ≤ 0.3* ≤ 0.6* 80–110
(< 27%)
HorRat ≤ 2
FB2 ≤ 0.15* ≤ 0.3* 80–110
(< 32%)

Note: *The LOD and LOQ were derived based upon typical B1:B2 ratio of 5:2 in naturally contaminated
samples
Maize Flour/Meal

Analyte ML (mg/Kg) LOD (mg/Kg) LOQ (mg/Kg) RSDR Recovery (%)

FB1 + FB2 2.0 - - - -
HorRat ≤ 2
FB1 ≤ 0.15* ≤ 0.3* 80–110
(< 30%)
HorRat ≤ 2
FB2 ≤ 0.06* ≤ 0.15* 80–110
(< 34%)

Note: *The LOD and LOQ were derived based upon typical B1:B2 ratio of 5:2 in naturally contaminated samples
CXS 193-1995 53

OCHRATOXIN A
Reference to JECFA: 37 (1990), 44 (1995), 56 (2001), 68 (2007)
Toxicological guidance value: PTWI 0.0001 mg/kg bw (2001)
Contaminant definition: Ochratoxin A
Synonyms: (The term “ochratoxins” includes a number of related mycotoxins (A, B, C and their esters and metabolites), the
most important one being ochratoxin A)
Related code of practice: Code of practice for the prevention and reduction of mycotoxin contamination in cereals (CXC 51-2003)11
Code of practice for the prevention and reduction of ochratoxin a contamination in wine (CXC 63-2007)12
Code of practice for the prevention and reduction of ochratoxin a contamination in coffee (CXC 69-2009)13
Code of practice for the prevention and reduction of ochratoxin a contamination in cocoa (CXC 72-2013)14
Code of practice for the prevention and reduction of mycotoxins in spices (CXC 78-2017)10
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name µg/kg product to which the ML applies
The ML applies to raw common wheat, raw durum wheat, raw spelt and
Wheat 5 Whole commodity
raw emmer.
Barley 5 Whole commodity The ML applies to raw barley.
Rye 5 Whole commodity The ML applies to raw rye.
The relevant Codex commodity standards are CXS 352-2022 and
Chili pepper, paprika, CXS 353-2022.
20 Whole/Powder/Crushed/Ground
nutmeg
The ML applies to spices (dried/dry).
CXS 193-1995 54

PATULIN
Reference to JECFA: 35 (1989), 44 (1995)
Toxicological guidance value: PMTDI 0.0004 mg/kg bw (1995)
Contaminant definition: Patulin
Related code of practice: Code of practice for the prevention and reduction of patulin contamination in apple juice and apple juice
ingredients in other beverages (CXC 50-2003)15
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name µg/kg product to which the ML applies
The relevant Codex commodity standard is CXS 247-2005
Whole commodity (not concentrated) (apple products only).
Apple juice 50 or commodity reconstituted to the
original juice concentration The ML applies also to apple juice used as an ingredient in other
beverages.
CXS 193-1995 55

ARSENIC
Reference to JECFA: 5 (1960), 10 (1967), 27 (1983), 33 (1988), 72 (2010)
Toxicological guidance value: At the 72nd meeting of JECFA (2010), the inorganic arsenic lower limit on the benchmark dose for a 0.5% increased
incidence of lung cancer (BMDL 0.5) was determined from epidemiological studies to be 3.0 μg/kg bw/day (2–
7 μg/kg bw/day based on the range of estimated total dietary exposure) using a range of assumptions to estimate
total dietary exposure to inorganic arsenic from drinking water and food. The JECFA noted that the provisional
tolerable weekly intake (PTWI) of 15 μg/kg bw (equivalent to 2.1 μg/kg bw/day) is in the region of the BMDL 0.5
and, therefore, was no longer appropriate. The JECFA withdrew the previous PTWI.
Contaminant definition: Arsenic: total (As-tot) when not otherwise mentioned; inorganic arsenic (As-in); or other specification
Synonyms: As
Related code of practice: Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Code of practice for the prevention and reduction of arsenic contamination in rice (CXC 77-2017)17
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
The relevant Codex commodity standards are CXS 19-1981,
CXS 33-1981, CXS 210-1999, CXS 211-1999 and CXS 329-2017.
For fish oils covered by CXS 329-2017, the ML is for fish oils (As-in).
Countries or importers may decide to use their own screening when
Edible fats and oils 0.1 Whole commodity applying the ML for As-in in fish oils by analysing total arsenic (As-tot) in
fish oils. If the As-tot concentration is below the ML for As-in, no further
testing is required, and the sample is determined to be compliant with
the ML. If the As-tot concentration is above the ML for As-in, follow-up
testing shall be conducted to determine if the As-in concentration is
above the ML.
Fat spreads and
0.1 The relevant Codex commodity standard is CXS 256-2007.
blended spreads
Natural mineral The relevant Codex commodity standard is CXS 108-1981.
0.01
waters Calculated as total As in mg/l.
The ML is for inorganic arsenic (As-in).
Countries or importers may decide to use their own screening when
applying the ML for As-in in rice by analysing total arsenic (As-tot) in
Rice, husked 0.35 Whole commodity rice. If the As-tot concentration is below or equal to the ML for As-in, no
further testing is required, and the sample is determined to be compliant
with the ML. If the As-tot concentration is above the ML for As-in, follow-
up testing shall be conducted to determine if the As-in concentration is
above the ML.
CXS 193-1995 56

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies
The ML is for inorganic arsenic (As-in).
Countries or importers may decide to use their own screening when
applying the ML for As-in in rice by analysing total arsenic (As-tot) in
Rice, polished 0.2 Whole commodity rice. If the As-tot concentration is below or equal to the ML for As-in, no
further testing is required, and the sample is determined to be compliant
with the ML. If the As-tot concentration is above the ML for As-in, follow-
up testing shall be conducted to determine if the As-in concentration is
above the ML.
Salt, food grade 0.5 The relevant Codex commodity standard is CXS 150-1985.
CXS 193-1995 57

CADMIUM
Reference to JECFA: 16 (1972), 33 (1988), 41 (1993), 55 (2000), 61 (2003), 64 (2005), 73 (2010), 77 (2013), 91(2021)
Toxicological guidance value: In view of the long half-life of cadmium, daily ingestion in food has a small or even a negligible effect on overall
exposure. In order to assess long- or short-term risks to health due to cadmium exposure, dietary intake should be
assessed over months, and tolerable intake should be assessed over a period of at least 1 month. To encourage
this view, at the 73rd meeting (2010) the JECFA decided to express the tolerable intake as a monthly value in the
form of a provisional tolerable monthly intake (PTMI) and established a PTMI of 25 μg/kg bw.
Contaminant definition: Cadmium, total
Synonyms: Cd
Related code of practice: Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Code of practice for the prevention and reduction of cadmium in cocoa beans (CXC 81-2022)18

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies

Head cabbages and kohlrabi: whole

commodity as marketed, after
removal of obviously decomposed or
withered leaves.
Brassica vegetables 0.05 The ML does not apply to Brassica leafy vegetables.
Cauliflower and broccoli: flower
heads (immature inflorescence
only).
Brussels sprouts: “buttons” only.
Bulb/dry onions and garlic: whole
commodity after removal of roots
Bulb vegetables 0.05
and adhering soil and whatever
parchment skin is easily detached.
Whole commodity after removal of
stems.
Fruiting vegetables 0.05 The ML does not apply to tomatoes and edible fungi.
Sweet corn and fresh corn: kernels
plus cob without husk.
Whole commodity as usually
marketed, after removal of obviously
Leafy vegetables 0.2 decomposed or withered leaves. The ML also applies to Brassica leafy vegetables.
CXS 193-1995 58

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies

Whole commodity as consumed.

The succulent forms may be
Legume vegetables 0.1
consumed as whole pods or as the
shelled product.
Pulses 0.1 Whole commodity The ML does not apply to soybeans (dry).
Whole commodity after removing
tops. Remove adhering soil (e.g. by
Root and tuber rinsing in running water or by gentle
0.1 The ML does not apply to celeriac.
vegetables brushing of the dry commodity).
Potato: peeled potato.
Whole commodity as marketed after
removal of obviously decomposed or
withered leaves.
Stalk and stem Rhubarb: leaf stems only.
0.1
vegetables
Globe artichoke: flower head only.
Celery and asparagus: remove
adhering soil.

Cereal grains 0.1 Whole commodity The ML does not apply to buckwheat, cañihua, quinoa, wheat and rice.

Rice, polished 0.4 Whole commodity

Wheat 0.2 Whole commodity The ML applies to common wheat, durum wheat, spelt and emmer.
Quinoa 0.15 Whole commodity The relevant Codex commodity standard is CXS 333-2019.

Marine bivalve Whole commodity after removal of The ML applies to clams, cockles and mussels but not to oysters and
2
molluscs shell scallops.

Whole commodity after removal of

Cephalopods 2 The ML applies to cuttlefishes, octopuses and squids without viscera.
shell
Natural mineral The relevant Codex commodity standard is CXS 108-1981.
0.003
waters The ML is expressed in mg/l.
Salt, food grade 0.5 The relevant Codex commodity standard is CXS 150-1985.
CXS 193-1995 59

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies

Chocolates containing
Including milk chocolate, family milk chocolate, milk chocolate
or declaring < 30% Whole commodity as prepared for
0.3 couverture, Gianduja milk chocolate, table chocolate, milk chocolate
total cocoa solids on a wholesale or retail distribution
Vermicelli/milk chocolate flakes.
dry matter basis

Chocolate containing
or declaring ≥ 30% to Including sweet chocolate, Gianduja chocolate, semi-bitter table
Whole commodity as prepared for
< 50% total cocoa 0.7 chocolate, Vermicelli chocolate/chocolate flakes, bitter table chocolate,
wholesale or retail distribution
solids on a dry matter couverture chocolate.
basis
Chocolate containing
or declaring ≥ 50% to Including sweet chocolate, Gianduja chocolate, semi-bitter table
Whole commodity as prepared for
< 70% total cocoa 0.8 chocolate, Vermicelli chocolate/chocolate flakes, and bitter table
wholesale or retail distribution
solids on a dry matter chocolate.
basis
Chocolate containing
Including sweet chocolate, Gianduja chocolate, semi-bitter table
or declaring ≥ 70% Whole commodity as prepared for
0.9 chocolate, Vermicelli chocolate/chocolate flakes, and bitter table
total cocoa solids on a wholesale or retail distribution
chocolate.
dry matter basis
Cocoa powder (100% The ML applies to 100% cocoa powder. The ML applies also to cocoa
total cocoa solids on a Whole commodity as prepared for powder when used as an ingredient in other foods. The ML does not
2.0
dry matter basis) wholesale or retail distribution apply to cocoa powder-based drink mixes that contain other ingredients,
ready for consumption such as milk powder and sugar.
CXS 193-1995 60

LEAD
Reference to JECFA: 10 (1966), 16 (1972), 22 (1978), 30 (1986), 41 (1993), 53 (1999), 73 (2010)
Toxicological guidance value: Based on the dose–response analyses, at the 73rd meeting (2010), JECFA estimated that the previously
established PTWI of 25 μg/kg bw is associated with a decrease of at least three intelligence quotient (IQ) points in
children and an increase in systolic blood pressure of approximately 3 mmHg (0.4 kPa) in adults. While such effects
may be insignificant at the individual level, these changes are important when viewed as a shift in the distribution
of IQ or blood pressure within a population. The JECFA, therefore, concluded that the PTWI could no longer be
considered health-protective and withdrew it.
Contaminant definition: Lead, total
Synonyms: Pb
Related code of practice: Code of practice for the prevention and reduction of lead contamination in foods (CXC 56-2004)19
Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
Berries and other Whole commodity after removal of caps and
0.1 The ML does not apply to cranberry, currant and elderberry.
small fruits stems.
Whole commodity after removal of caps and
Cranberry 0.2
stems.

Currants 0.2 Fruit with stem.

Whole commodity after removal of caps and

Elderberry 0.2
stems.
Whole commodity.
Berries and other small fruits: whole
commodity after removal of caps and stems.
Pome fruits: whole commodity after
removal of stems.
Fruits 0.1 Stone fruits, dates, and olives: whole The ML does not apply to cranberry, currant and elderberry.
commodity after removal of stems and
stones, but the level calculated and
expressed on the whole commodity without
stem.
Pineapple: whole commodity after
removal of crown.
CXS 193-1995 61

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies
Avocado, mangos, and similar fruit with hard
seeds: whole commodity after removal of
stone but calculated on whole fruit.
Head cabbages and kohlrabi: whole
commodity as marketed, after removal of
obviously decomposed or withered leaves.
Brassica vegetables 0.1 The ML does not apply to kale and leafy Brassica vegetables.
Cauliflower and broccoli: flower heads
(immature inflorescence only).
Brussels sprouts: “buttons” only.
Bulb/dry onions and garlic: whole commodity
Bulb vegetables 0.1 after removal of roots and adhering soil and
whatever parchment skin is easily detached.
Whole commodity after removal of stems
Fruiting vegetables 0.05 Sweet corn and fresh corn: kernels plus cob The ML does not apply to fungi and mushrooms.
without husk.
Whole commodity as usually marketed, after
The ML applies to leafy Brassica vegetables but does not apply
Leafy vegetables 0.3 removal of obviously decomposed or
to spinach.
withered leaves.
Whole commodity as consumed. The
Legume vegetables 0.1 succulent forms may be consumed as whole
pods or as the shelled product.
Fresh farmed
mushrooms (common
mushrooms (Agaricus
bisporous), shiitake
0.3 Whole commodity The relevant Codex commodity standard is CXS 38-1981.
mushrooms (Lentinula
edodes), and oyster
mushrooms
(Pleurotus ostreatus))
Pulses 0.1 Whole commodity
Whole commodity after removing tops.
Remove adhering soil (e.g. by rinsing in
Root and tuber running water or by gentle brushing of the
0.1
vegetables dry commodity).
Potato: peeled potato.
CXS 193-1995 62

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies
The relevant Codex commodity standards are CXS 242-2003,
Canned fruits 0.1 The ML applies to the product as consumed. CXS 254-2007, CXS 78-1981, CXS 159-1987, CXS 42-1981,
CXS 99-1981, CXS 60-1981, CXS 62-1981.
Jams, jellies and The relevant Codex commodity standard is CXS 296-2009 (for
0.4
marmalades jams and jellies only).
Mango chutney 0.4 The relevant Codex commodity standard is CXS 160-1987.
Canned vegetables 0.1 The ML applies to the product as consumed The relevant Codex commodity standard is CXS 297-2009.
The relevant Codex commodity standard is CXS 13-1981.
In order to consider the concentration of the product, the
Preserved tomatoes 0.05 determination of the maximum levels for contaminants shall
consider the natural total soluble solids, the reference value
being 4.5 for fresh fruit.
Table olives 0.4 The relevant Codex commodity standard is CXS 66-1981.
Pickled cucumbers
0.1 The relevant Codex commodity standard is CXS 115-1981.
(cucumber pickles)
Canned chestnuts
and canned chestnuts 0.05 The relevant Codex commodity standard is CXS 145-1985.
puree
The ML does not apply to juices exclusively from berries and
Whole commodity (not concentrated) or other small fruit.
commodity reconstituted to the original juice
Fruit juices 0.03 The relevant Codex commodity standard is CXS 247-2005.
concentration, ready-to-drink. The ML also
applies to nectars, ready-to-drink. The ML also applies to fruit juices for infants and young
children.
Fruit juices obtained Whole commodity (not concentrated) or
exclusively from commodity reconstituted to the original juice The ML does not apply to grape juice.
0.05
berries and other concentration, ready-to-drink. The ML also The relevant Codex commodity standard is CXS 247-2005.
small fruits applies to nectars, ready-to-drink.
Whole commodity (not concentrated) or The relevant Codex commodity standard is CXS 247-2005.
commodity reconstituted to the original juice
Grape juice 0.04 The ML also applies to fruit juices for infants and young
concentration, ready-to-drink. The ML also
applies to nectars, ready-to-drink. children.

Cereal grains 0.2 Whole commodity The ML does not apply to buckwheat cañihua and quinoa.
CXS 193-1995 63

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies

Cereal-based foods The relevant Codex commodity standard is CXS 74-1981.

Whole commodity as sold; not reconstituted
for infants and young 0.02 The ML applies to all cereal-based foods intended for infants
or otherwise prepared for consumption
children (up to 12 months) and young children (12 to 36 months).

Infant formula,
formula for special The relevant Codex commodity standards are CXS 72-1981
medical purposes 0.01 Whole commodity and CXS 156-1987.
intended for infants The ML applies to the formula as consumed.
and follow-up formula
Ready-to-eat meals The ML applies to all ready-to-eat meals intended for infants
for infants and young 0.02 Whole commodity (up to 12 months) and young children (12 to 36 months). The
children relevant Codex commodity standard is CXS 73-1981.
Quinoa 0.2 Whole commodity The relevant Codex commodity standard is CXS 333-2019.
Whole commodity
Fish 0.3
(in general after removing the digestive tract)
Meat of cattle, pigs
0.1 Whole commodity (without bones) The ML also applies to fat from the meat.
and sheep
Meat and fat of
0.1 Whole commodity (without bones)
poultry
Edible offal means such offal as have been passed as fit for
human consumption, but not including lungs, ears, scalp, snout
(including lips and muzzle), mucous membranes, sinews,
Cattle, edible offal of 0.2 Whole commodity genital system, udders, intestines, and urinary bladder
(CXA 4-1989).
The ML applies to the following edible offal: Brain, head, heart,
kidney, liver, tongue and stomach.
Edible offal means such offal as have been passed as fit for
human consumption, but not including lungs, ears, scalp, snout
(including lips and muzzle), mucous membranes, sinews,
Pig, edible offal of 0.15 Whole commodity genital system, udders, intestines, and urinary bladder
(CXA 4-1989).
The ML applies to the following edible offal: Blood, heart,
kidney, liver and tongue.
CXS 193-1995 64

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies
Poultry edible offal are such edible tissues and organs, other
than poultry meat and poultry fat, from slaughtered poultry as
Poultry, edible offal of 0.1 Whole commodity have been passed fit for human consumption (CXA 4-1989).
The ML applies to the following edible offal: Heart, kidney, liver,
stomach and thymus.
The relevant Codex commodity standards are CXS 19-1981,
Whole commodity as prepared for wholesale
Edible fats and oils 0.08 CXS 33-1981, CXS 210-1999, CXS 211-1999 and
or retail distribution
CXS 329-2017.
Fat spreads and Whole commodity as prepared for wholesale
0.04 The relevant Codex commodity standard is CXS 256-2007.
blended spreads or retail distribution
Milk is the normal mammary secretion of milking animals
obtained from one or more milkings without either addition to it
or extraction from it, intended for consumption as liquid milk or
Milk 0.02 Whole commodity for further processing.
A concentration factor applies to partially or wholly dehydrated
milks.
Secondary milk
0.02 Whole commodity The ML applies to the food as consumed.
products
Natural mineral The relevant Codex commodity standard is CXS 108-1981.
0.01
waters The ML is expressed in mg/l.
Whole commodity as prepared for wholesale The relevant Codex commodity standard is CXS 150-1985.
Salt, food grade 1
or retail distribution Excluding salt from marshes.
The ML applies to soft brown sugar, raw sugar and non-
Soft brown, raw and
0.15 Whole commodity centrifugal sugar.
non-centrifugal sugars
The relevant Codex commodity standard is CXS 212-1999.
White and refined
The relevant Codex commodity standards are CXS 212-1999
sugar, corn and 0.1 Whole commodity
(white and refined sugar) and CXS 12-1981 (honey).
maple syrups, honey
Candies, sugar-based 0.1 Whole commodity The ML applies all sugar-based candies.
Spices, dried aril 0.9 Whole, ground, powder, crushed
Spices, dried floral
2.5 Whole, ground, powder, crushed The relevant Codex commodity standard is CXS 344-2021.
parts
CXS 193-1995 65

Commodity/ Maximum level (ML) Portion of the commodity/

Notes/Remarks
Product name mg/kg product to which the ML applies
Spices, dried fruit and The ML does not apply to dried Sichuan pepper, star anise,
0.6 Whole, ground, powder, crushed
berries dried paprika and sumac.
Spices, dried paprika
0.8 Whole, ground, powder, crushed The relevant Codex commodity standard is CXS 353-2022.
and sumac
Spices, dried Sichuan
3.0 Whole, ground, powder, crushed
pepper and star anise
Spices, dried
2.0 Whole, ground, powder, crushed The relevant Codex commodity standard is CXS 343-2021.
rhizomes and roots
The ML does not apply to dried celery seeds. The relevant
Spices, dried seeds 0.9 Whole, ground, powder, crushed Codex commodity standards are CXS 327-2021 and
CXS 352-2022.
Spices, dried celery
1.5 Whole, ground, powder, crushed
seeds
The ML applies to wine made from grapes harvested after the
Wine 0.1 Whole commodity
date of adoption (CAC42, July 2019).
Wine (wine and The ML applies to wines and fortified/liqueur wines made from
0.2 Whole commodity
fortified/liqueur wine) grapes harvested before (CAC42, July 2019).
The ML applies to wine made from grapes harvested after the
Fortified/Liqueur wine 0.15 Whole commodity
date of adoption (CAC42, July 2019).
CXS 193-1995 66

MERCURY
Reference to JECFA: 10 (1966), 14 (1970), 16 (1972), 22 (1978), 72 (2010)
Toxicological guidance value: At the 72nd meeting (2010), JECFA established a PTWI for inorganic mercury of 4 μg/kg bw. The previous PTWI
of 5 μg/kg bw for total mercury, established at the sixteenth meeting, was withdrawn. The new PTWI for inorganic
mercury was considered applicable to dietary exposure to total mercury from foods other than fish and shellfish.
For dietary exposure to mercury from these foods, the previously established PTWI for methyl mercury should be
applied.
Contaminant definition: Mercury, Total
Synonyms: Hg
Related code of practice: Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
Natural mineral The relevant Codex commodity standard is CXS 108-1981.
0.001
waters The ML is expressed in mg/l.
Salt food grade 0.1 The relevant Codex commodity standard is CXS 150-1985.
CXS 193-1995 67

METHYLMERCURY
Reference to JECFA: 22 (1978), 33 (1988), 53 (1999), 61 (2003), 67 (2006)
Toxicological guidance value: PTWI 0.0016 mg/kg bw (2003, confirmed in 2006)
Contaminant definition: Methylmercury
Related code of practice: Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name (mg/kg) product to which the ML applies
Countries or importers may decide to use their own screening when
Alfonsino 1.5 applying the ML for methylmercury in fish by analysing total mercury
in fish. If the total mercury concentration is below or equal to the ML
Marlin 1.7 for methylmercury, no further testing is required, and the sample is
determined to be compliant with the ML. If the total mercury
concentration is above the ML for methylmercury, follow-up testing
Orange roughy 0.8 Whole commodity fresh or frozen shall be conducted to determine if the methylmercury concentration is
(in general, after removing the above the ML.
digestive tract) The ML also applies to fresh or frozen fish intended for further
Pink cusk-eel 1.0
processing.
Countries should consider developing nationally relevant consumer
Shark 1.6 advice for women of childbearing age and young children to
supplement the ML.
Tuna 1.2 For sampling plan, see Annex IX.
CXS 193-1995 68

Annex IX
SAMPLING PLAN FOR METHYLMERCURY CONTAMINATION IN FISH
GENERAL CONSIDERATIONS
DEFINITION

An identifiable quantity of a food commodity delivered at one time and

determined by the official to have common characteristics such as origin,
variety, type of packing, packer, consignor or markings.
Lot A lot of whole fish should consist of one species and the length and/or weight
should be comparable. In case the length and/or weight of the fish is not
comparable, the consignment may still be considered as a lot, but a specific
sampling procedure has to be applied (as described in paragraph 8).

Designated part of a larger lot in order to apply the sampling method on that
Sublot
designated part. Each sublot must be physically separate and identifiable.

A procedure for sampling of food from a certain lot with a view of a specific
chemical analysis of that lot, in order to ensure that the sample that is taken,
Sampling plan
is representative for the concentration of the concerned chemical within the
lot.

A methylmercury test procedure consists of three steps: sample selection,

Methylmercury test
sample preparation and methylmercury quantification. It contains an
procedure
accept/reject level.

The lot is accepted if the test result is less than or equal to the Codex
Decision rule
maximum level (ML); otherwise, the lot is rejected.

Incremental sample The quantity of material taken from a single random place in the lot or sublot.

The combined total of all the incremental samples that is taken from the lot or
sublot. The aggregate sample has to be at least as large as the laboratory
Aggregate sample
sample or samples combined. The entire aggregate sample should be
comminuted in a mill.

A sample intended for the laboratory, which consists out of a comminuted

quantity of fish muscle, or whole fish. The laboratory sample may be a portion
Laboratory sample of or the entire aggregate sample. If the aggregate sample is larger than the
laboratory sample/s, the laboratory sample(s) should be removed in a
random manner from the homogenized aggregate sample.

A randomly removed portion of the comminuted laboratory sample for the

Test portion
extraction of the methylmercury for chemical analysis.

MATERIAL TO BE SAMPLED
1. Each lot or sublot which is to be examined must be sampled separately.
2. Fresh or frozen whole (in general after removing digestive tract) or dressed fish (eviscerated fish with
head and tail removed) and other non-bulk fishery products of lots greater than or equal to 15 metric
tonnes (MT) should be subdivided into sublots of 15–30 MT in accordance with Table 2.
3. Lots of fishery products traded as bulk commodities of greater than 100 MT should be subdivided into
sublots in accordance with Table 1 to be sampled separately.
CXS 193-1995 69

Table 1. Subdivision of sublots according to bulk consignment lot weight

Weight or number of
Commodity Lot weight (MT*)
sublots (MT)

≥ 1 500 500

3 sublots
Fishery products > 300 and < 1 500
(minimum 100 MT)
(traded as bulk
consignments)
≥ 100 and ≤ 300 100

< 100 -

Note: *1 metric tonne (MT) = 1 000 kilograms

Table 2. Subdivision of sublots according to other products lot weight

Weight or number of
Commodity Lot weight (MT*)
sublots (MT)

Fish (traded as ≥ 15 15 – 30
non-bulk
consignments) < 15 -

Note: *1 metric tonne (MT) = 1 000 kilograms

4. Taking into account that the weight of the lot is not always an exact multiple of the weight of the sublots,
the weight of the sublot may exceed the mentioned weight by a maximum of 20 percent
INCREMENTAL SAMPLE
5. The recommended minimum number of incremental samples taken from the lot or sublot is dependent
on the size of the lot or sublot as specified in Table 3.
6. The aggregate sample should contain a quantity of sample of at least 1 kilogram. The minimum weight
of the incremental sample should be determined by dividing 1 kilogram by the required number of
incremental samples as listed in Table 3. Incremental samples taken from a lot or sublot should be of
comparable weight.
Table 3. Number of incremental samples to be taken depending on the weight of the lot or sublot

Minimum
Number of
Lot weight (MT*) laboratory sample
incremental samples
weight (kg)

≤ 0.05 3 1

> 0.05 – ≤ 0.5 5 1

> 0.5 10 1
Note: *1 metric tonne (MT) = 1 000 kilograms
7. Whole fish are considered to be of comparable length and weight class where the differences in size
and/or weight do not exceed about 50 percent.
8. For lots where fish are not of comparable length and/or weight the following approaches are to be
applied to taking the incremental samples:
a. Where a length or weight class/category is predominant (80 percent or more of the fish lot or
sublot are within the same length and/or weight class), the aggregate sample is combined only
from incremental samples of fish within the predominant category and outliers are excluded. This
aggregate sample is to be considered as being representative for the whole lot/sublot.
b. Where there is no predominant weight or size class and where the overall length and/or weight
of the fish present in the lot or sublot varies by more than 50 percent but less than 100 percent,
the lot or sublot is separated into two length or weight classes and separate aggregate samples
are composited from incremental samples taken independently from each length and/or weight
class.
CXS 193-1995 70

c. Where there is no predominant weight or size class and where the overall length and/or weight
of the fish present in the lot differ more than 100 percent, the lot or sublot is separated into three
length or weight classes and separate aggregate samples are composited from incremental
samples taken independently from each length or weight class.
9. For lots or sublots of whole fish the part of the fish where the incremental sample is taken is informed
by the weight of the whole fish as specified in Table 3. Some examples on sampling of batches of fishes
of different size and/or weight can be found in Annex IX.A.
Table 4. Tissue area the incremental sample is taken from for whole fish based on weight classes

Weight class of an individual whole fish Sampled part

Whole fish (after removing the digestive tract)

For lots of 0.05 MT* or greater where the aggregate
< 1 kg sample would exceed 3 kg the midline (halfway
between the gill opening and the anus) strip from
backbone to belly can be sampled

Midline (halfway between the gill opening and the

1 – 6 kg
anus) strip from backbone to belly

Midline (halfway between the gill opening and the

anus) strip from backbone to belly
Alternatively, equal composite parts of muscle from
> 6 kg
behind the head and close to the tail can be sampled
For tuna, incremental samples can instead be taken
from the muscle from close to the tail

Note: *1 metric tonne (MT) = 1 000 kilograms

PACKAGING AND TRANSPORTATION OF SAMPLES

10. Each laboratory sample should be placed in a clean, inert container offering adequate protection from
contamination, loss of analytes by adsorption to the internal wall of the container and against damage
in transit. All necessary precautions, for example temperature control and storage in airtight containers,
should be taken to avoid any change in composition of the sample which might arise during
transportation or storage (for example avoiding excess heat or the sample drying out).
SEALING AND LABELLING OF SAMPLES
11. Each laboratory sample taken for official use shall be sealed at the place of sampling and identified. A
record must be kept of each sample, permitting each lot or sublot, to be clearly identified and giving the
date and place the sampling occurred, together with any additional information likely to be of assistance
to the analyst.
SAMPLE PREPARATION PRECAUTIONS
12. In the course of sampling, precautions, such as correct sampling technique and limitation of cross-
contamination, should be taken to avoid any changes which would affect the levels of methylmercury,
adversely affect the analytical determination, or make the aggregate samples unrepresentative.
13. Wherever possible, apparatus and equipment coming into contact with the sample should not contain
mercury and should be made of inert materials, e.g. plastics such as polypropylene,
polytetrafluoroethylene (PTFE) etc. These should be acid cleaned to minimize the risk of contamination.
High quality stainless steel may be used for cutting edges to take increment samples and make
comminuted samples.
HOMOGENIZATION – GRINDING
14. The complete aggregate sample should be finely comminuted and thoroughly mixed using a process
that has been demonstrated to achieve complete homogenization. Depending on the equipment
available frozen samples may need to be thawed prior to homogenization.
CXS 193-1995 71
CXS 193-1995 72

TEST PORTION
15. Procedures for selecting the test portion from the comminuted laboratory sample should be a random
process. Following homogenization and thorough mixing, the test portion can be selected from any
location throughout the comminuted laboratory sample.
16. It is suggested that three test portions be selected from each comminuted laboratory sample. The three
test portions will be used for enforcement, appeal and confirmation if needed.
ANALYTICAL METHODS
17. A criteria-based approach, whereby a set of performance criteria is established with which the analytical
method used should comply, is appropriate. The performance criteria-based approach has the
advantage that, by avoiding setting down specific details of the method used, developments in
methodology can be exploited without having to reconsider or modify the specific method. Utilizing this
approach, laboratories would be free to use the analytical method most appropriate for their facilities.
18. Refer to the Procedural Manual of the Codex Alimentarius Commission 1 for principles for the
establishment of methods of analysis.
19. Method performance criteria for methylmercury and total mercury are detailed for the species of fish for
which there are Codex MLs in the Recommended methods of analysis and sampling (CXS 234-1999).20
20. Countries or importers may decide to use their own screening when applying the ML for methylmercury
in fish by analysing total mercury in fish. If the total mercury concentration is below or equal to the ML
for methylmercury, no further testing is required, and the sample is determined to be compliant with the
ML. If the total mercury concentration is above the ML for methylmercury, follow-up testing shall be
conducted to determine if the methylmercury concentration is above the ML.
RECONDITIONING LOTS/SUBLOTS
21. A lot or sublot where fish are not of comparable length and/or weight that is separated into 2 to 3 length
and/or weight classes should be analysed sequentially from the largest class first.
22. A lot or sublot where fish are not of comparable length and/or weight can be considered in compliance
with the ML if the methylmercury concentration of the aggregate sample taken from the highest length
and/or weight class is below the ML. However, export or trade requirements (e.g. certificates of analysis)
may require testing lots or sublots of smaller length and/or weight classes.
23. Where the methylmercury concentration in the aggregate sample taken from a length and/or weight
class is above the ML then the next largest length/weight class should also be analysed. If the
methylmercury concentration in this sample is below the ML the lot or sublot can be reconditioned to
remove length and/or weight classes that exceed the ML to ensure the remaining fish are in compliance
with the ML.
24. For a lot or sublot separated into three length or weight classes paragraph 23 should be repeated for
the smallest length/weight classes if the methylmercury concentration in the aggregate sample taken
from the middle length/weight class is also above the ML.
CXS 193-1995 73

Annex IX.A
Examples on how to apply provisions in the sampling plan
EXAMPLE 1
In case the size and/or weight of the fish present in the lot differs more than 50 percent but less than
100 percent: two separate representative samples are taken from each size or weight class/category
within a lot.
Example: 5 MT (metric tonnes) lot of fishes with weights from 2 kg to 3.5 kg.
A first aggregate sample is taken of the smaller sized (lot relative) fishes, which weigh about 2–2.75 kg: ten
incremental samples (fishes) are taken. Each incremental sample is constituted from the muscle meat of the
middle part of the fish (slice backbone to belly, symmetrically taken around line B in Figure 1) and weighs
about 100 grams. This results in one aggregate sample of about 1 kg to be homogenized and analysed
separately.
A second aggregate sample is taken of the larger-sized (lot relative) fishes, which weigh about 2.75–3.5 kg:
10 incremental samples (fishes) are taken. Each incremental sample is constituted from the muscle meat of
the middle part of the fish (slice backbone to belly, symmetrically taken around line B in Figure 1) and weighs
about 100 grams. This results in one aggregate sample of about 1 kg to be homogenized and analysed
separately.

Figure 1: The different sections of a fish

A B C

A) Laboratory performs a sequential analysis:

First, the sample of the larger-sized fishes is homogenized and analysed separately.
 In case the analytical result is compliant, the whole lot is compliant.
 In case the analytical result is non-compliant, as a second step the sample of the smaller sized fishes is
homogenized and analysed separately.
 In case the analytical result of the sample of the smaller sized fishes is non-compliant, the whole lot is
non-compliant.
 In case the analytical result of the sample of smaller sized fishes is compliant, then the smaller sized
fishes (2–2.75 kg) have to be sorted out and these fishes are compliant. The remaining larger-sized
fishes (2.75–3.5 kg) are non-compliant.
B) Laboratory analyses both samples at the same time:
 In case both analytical results are compliant, the whole lot is compliant.
 In case both analytical results are non-compliant, the whole lot is non-compliant.
 In case the sample of the smaller sized fishes (2–2.75 kg) is compliant and the sample of the larger-
sized fishes (2.75-3.5 kg) not, then the smaller sized fishes (2–2.75 kg) have to be sorted out and these
small sized fishes are compliant. The remaining larger-sized fish (2.75–3.5 kg) are non-compliant.
CXS 193-1995 74

EXAMPLE 2
In case the size and/or weight of the fishes present in the lot differs more than 100 percent: three
separate representative samples are taken from each size or weight class/category within a lot
Example: 10 MT (metric tonnes) lot of fishes with weights from 2 kg to 8 kg.
A first aggregate sample is taken of the smaller sized (lot relative) fishes, which weigh about 2-
4 kg: 10 incremental samples (fishes) are taken, each incremental sample is constituted from the muscle meat
of the middle part of the fish (slice backbone to belly, symmetrically taken around line B in Figure 1) and weighs
about 100 grams. This results in one aggregate sample of about 1 kg, to be homogenized and analysed
separately.
A second aggregate sample is taken of the fishes of medium size (lot relative) of about 4–6 kg: ten incremental
samples (fishes) are taken, each incremental sample is constituted from the muscle meat of the middle part of
the fish (slice backbone to belly) and weighs about 100 grams. This results in one aggregate sample of about
1 kg, to be homogenized and analysed separately.
A third aggregate sample is taken of the larger-sized (lot relative) fishes of about 6–8 kg: ten incremental
samples (fishes) are taken, each incremental sample is
 constituted of the right side dorso-lateral muscle meat in the middle part of the fish (symmetrically around
line B in Figure 1 and above the horizontal line in Figure 1) and weighs about 100 grams. This results
in one aggregate sample of about 1 kg to be homogenized and analysed separately.
OR
 constituted of equal parts of 50 grams of the muscled meat close to the tail part (the region around line C
in Figure 1) and the muscle meat close to the head part of one fish (the region of line A in Figure 1)
which are combined to form an incremental sample of about 100 grams per fish. This results in one
aggregate sample of about 1 kg to be homogenized and analysed separately.
CXS 193-1995 75

TIN
Reference to JECFA: 10 (1966), 14 (1970), 15 (1971), 19 (1975), 22 (1978), 26 (1982), 33 (1988), 55 (2000), 64 (2005)
Toxicological guidance value: PTWI 14 mg/kg bw (1988, expressed as Sn; includes tin from food additive uses; maintained in 2000)
Contaminant definition: Tin, total (Sn-tot) when not otherwise mentioned; inorganic tin (Sn-in); or other specification
Synonyms: Sn
Related code of practice: Code of practice for the prevention and reduction of inorganic tin contamination in canned foods (CXC 60-2005)21
Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
The ML does not apply to non-tinplate canned cooked cured chopped
meat, cooked cured ham, cooked cured pork shoulder, corned beef and
luncheon meat.
Canned foods (other The relevant Codex commodity standards include CXS 62-1981,
250 CXS 254-2007, CXS 296-2009, CXS 242-2003, CXS 297-2009,
than beverages)
CXS 78-1981, CXS 159-1987, CXS 42-1981, CXS 60-1981,
CXS 99-1981, CXS 160-1987, CXS 66-1981, CXS 13-1981,
CXS 115-1981, CXS 57-1981, CXS 145-1981, CXS 98-1981,
CXS 96-1981, CXS 97-1981, CXS 88-1981, CXS 89-1981.
Canned beverages 150 The relevant Codex commodity standards include CXS 247-2005.
Cooked cured The ML applies to products in containers other than tinplate containers.
50
chopped meat The relevant Codex commodity standard is CXS 98-1981.
The ML applies to products in containers other than tinplate containers.
Cooked cured ham 50
The relevant Codex commodity standard is CXS 96-1981.
Cooked cured pork The ML applies to products in containers other than tinplate containers.
50
shoulder The relevant Codex commodity standard is CXS 97-1981.
The ML applies to products in containers other than tinplate containers.
Corned beef 50
The relevant Codex commodity standard is CXS 88-1981.
The ML applies to products in containers other than tinplate containers.
Luncheon meat 50
The relevant Codex commodity standard is CXS 89-1981.
CXS 193-1995 76

RADIONUCLIDES
TABLE 1
Portion of the
Commodity/ Guideline level (GL) commodity/
Representative radionuclides Notes/Remarks
Product name (Bq/kg) product to which the
GL applies
The GL applies to foods intended for
Infant foods 1 Pu-238, Pu-239, Pu-240, Am-241
consumption by infants.
The GL applies to foods intended for
Infant foods 100 Sr-90, Ru-106, I-129, I-131, U-235
consumption by infants.
S-35*, Co-60, Sr-89, Ru-103, Cs-134, The GL applies to foods intended for
Infant foods 1 000
Cs-137, Ce-144, Ir-192 consumption by infants.
The GL applies to foods intended for
Infant foods 1 000 H-3**, C-14, Tc-99
consumption by infants.
Foods other than infant
10 Pu-238, Pu-239, Pu-240, Am-241
foods
Foods other than infant
100 Sr-90, Ru-106, I-129, I-131, U-235
foods
Foods other than infant S-35*, Co-60, Sr-89, Ru-103, Cs-134,
1 000
foods Cs-137, Ce-144, Ir-192
Foods other than infant
10 000 H-3**, C-14, Tc-99
foods
Notes
*This represents the value for organically bound sulphur.
**This represents the value for organically bound tritium.
Scope: The guideline levels apply to radionuclides contained in foods destined for human consumption and traded internationally, which have been contaminated
following a nuclear or radiological emergency.i These guideline levels apply to food after reconstitution or as prepared for consumption, i.e. not to dried or concentrated
foods, and are based on an intervention exemption level of 1 mSv in a year.
Application: As far as generic radiological protection of food consumers is concerned, when radionuclide levels in food do not exceed the corresponding guideline
levels, the food should be considered as safe for human consumption. When the guideline levels are exceeded, national governments shall decide whether and under
what circumstances the food should be distributed within their territory or jurisdiction. National governments may wish to adopt different values for internal use within
their own territories where the assumptions concerning food distribution that have been made to derive the guideline levels may not apply, e.g. in the case of widespread
radioactive contamination. For foods that are consumed in small quantities, such as spices, that represent a small percentage of total diet and hence a small addition
to the total dose, the guideline levels may be increased by a factor of ten.

i For the purposes of this document, the term “emergency” includes both accidents and malevolent actions.
CXS 193-1995 77

Radionuclides: The guideline levels do not include all radionuclides. Radionuclides included are those important for uptake into the food chain; are usually contained
in nuclear installations or used as a radiation source in large enough quantities to be significant potential contributors to levels in foods and could be accidentally
released into the environment from typical installations or might be employed in malevolent actions. Radionuclides of natural origin are generally excluded from
consideration in this document.
In the Table, the radionuclides are grouped according to the guideline levels rounded logarithmically by orders of magnitude. Guideline levels are defined for two
separate categories “infant foods” and “other foods”. This is because, for a number of radionuclides, the sensitivity of infants could pose a problem. The guideline
levels have been checked against age-dependent ingestion dose coefficients defined as committed effective doses per unit intake for each radionuclide, which are
taken from the “International Basic Safety Standards” (IAEA, 1996).ii
Multiple radionuclides in foods: The guideline levels have been developed with the understanding that there is no need to add contributions from radionuclides in
different groups. Each group should be treated independently. However, the activity concentrations of each radionuclide within the same group should be added
together.iii

ii Food and Agriculture Organization of the United Nations, International Atomic Energy Agency, International Labour Office, OECD Nuclear Energy Agency, Pan American Health
Organization, World Health Organization (1996) International Basic Safety Standards for Protection against Ionizing Radiation and for the Safety of Radiation Sources, IAEA,
Vienna.
iii For example, if 134Cs and 137Cs are contaminants in food, the guideline level of 1 000 Bq/kg refers to the summed activity of both these radionuclides.
CXS 193-1995 78

Annex X
SCIENTIFIC JUSTIFICATION FOR THE GUIDELINE LEVELS FOR RADIONUCLIDES IN FOODS
CONTAMINATED FOLLOWING A NUCLEAR OR RADIOLOGICAL EMERGENCY
The guideline levels for radionuclides in foods and specifically the values presented in Table 1 above are
based on the following general radiological considerations and experience of application of the existing
international and national standards for control of radionuclides in food.
Significant improvements in the assessment of radiation doses resulting from the human intake of radioactive
substances have become available since the guideline levels were issued by the Codex Alimentarius
Commission in 1989i (CXG 5-1989).
Infants and adults: The levels of human exposure resulting from consumption of foods containing
radionuclides listed in Table 1 at the suggested guideline levels have been assessed both for infants and adults
and checked for compliance with the appropriate dose criterion.
In order to assess public exposure and the associated health risks from intake of radionuclides in food,
estimates of food consumption rates and ingestion dose coefficients are needed. It is assumed that 550 kg of
food is consumed by an adult in a year. The value of infant food and milk consumption during first year of life
used for infant dose calculation equal to 200 kg is based on contemporary human habit assessments. The
most conservative values of the radionuclide-specific and age-specific ingestion dose coefficients, i.e. relevant
to the chemical forms of radionuclides which are most absorbed from the gastro-intestinal tract and retained
in body tissues, are taken from the International Atomic Energy Agency (IAEA).
Radiological criterion: The appropriate radiological criterion, which has been used for comparison with the
dose assessment data below, is a generic intervention exemption level of around 1 mSv for individual annual
dose from radionuclides in major commodities, e.g. food, recommended by the International Commission on
Radiological Protection as safe for members of the public.
Naturally-occurring radionuclides: Radionuclides of natural origin are ubiquitous and as a consequence are
present in all foodstuffs to varying degrees. Radiation doses from the consumption of foodstuffs typically range
from a few tens to a few hundreds of microsieverts in a year. In essence, the doses from these radionuclides
when naturally present in the diet are unamenable to control; the resources that would be required to affect
exposures would be out of proportion to the benefits achieved for health. These radionuclides are excluded
from consideration in this document as they are not associated with emergencies.
One-year exposure assessment: It is conservatively assumed that during the first year after major
environmental radioactive contamination caused by a nuclear or radiological emergency, it might be difficult to
readily replace foods imported from contaminated regions with foods imported from unaffected areas.
According to FAO statistical data, the mean fraction of major foodstuff quantities imported by all the countries
worldwide is 0.1. The values in Table 1 as regards foods consumed by infants and the general population have
been derived to ensure that if a country continues to import major foods from areas contaminated with
radionuclides, the mean annual internal dose of its inhabitants will not exceed around 1 mSv (see Annex XI).
This conclusion might not apply for some radionuclides if the fraction of contaminated food is found to be higher
than 0.1, as might be the case for infants who have a diet essentially based on milk with little variety.
Long-term exposure assessment: Beyond one year after the emergency the fraction of contaminated food
placed on the market will generally decrease as a result of national restrictions (withdrawal from the market),
changes to other produce, agricultural countermeasures and decay.
Experience has shown that in the long term, the fraction of imported contaminated food will decrease by a
factor of a hundred or more. Specific food categories, e.g. wild forest products, may show persistent or even
increasing levels of contamination. Other categories of food may gradually be exempted from controls.
Nevertheless, it must be anticipated that it may take many years before levels of individual exposure as a result
of contaminated food could be qualified as negligible.

i The Codex Alimentarius Commission at its 18th Session (Geneva 1989) adopted Guideline levels for radionuclides
in foods following accidental nuclear contamination for use in international trade (CXG 5-1989) applicable for six
radionuclides (90Sr, 131I, 137Cs, 134Cs, 239Pu and 241Am) during one year after the nuclear accident.
CXS 193-1995 79

Annex XI
ASSESSMENT OF HUMAN INTERNAL EXPOSURE WHEN THE GUIDELINE LEVELS ARE APPLIED
For the purpose of assessment of the mean public exposure level in a country caused by the import of food
products from foreign areas with residual radioactivity, in implementing the present guideline levels the
following data should be used: annual food consumption rates for infants and adults, radionuclide- and age-
dependent ingestion dose coefficients and the import/production factors. When assessing the mean internal
dose in infants and adults, it is suggested that due to monitoring and inspection the radionuclide concentration
in imported foods does not exceed the present guideline levels. Using cautious assessment approach, it is
considered that all the foodstuffs imported from foreign areas with residual radioactivity are contaminated with
radionuclides at the present guideline levels.
Then, the mean internal dose of the public, E (mSv), due to annual consumption of imported foods containing
radionuclides can be estimated using the following formula:
E = GL(A) M(A) eing(A) IPF
where:
GL(A) is the guideline level (Bq/kg)
M(A) is the age-dependent mass of food consumed per year (kg)
eing(A) is the age-dependent ingestion dose coefficient (mSv/Bq)
IPF is the import/production factori (dimensionless)
Assessment results presented in Table 2 both for infants and adults demonstrate that for all the
20 radionuclides’ doses from consumption of imported foods during the first year after major radioactive
contamination do not exceed 1 mSv. It should be noted that the doses were calculated on the basis of a value
for the IPF equal to 0.1 and that this assumption may not always apply, in particular to infants who have a diet
essentially based on milk with little variety.
It should be noted that for 239Pu as well as for a number of other radionuclides, the dose estimate is
conservative. This is because elevated gastro-intestinal tract absorption factors and associated ingestion dose
coefficients are applied for the whole first year of life whereas this is valid mainly during suckling period recently
estimated by ICRP to be as average first six months of life. For the subsequent six months of the first year of
life the gut absorption factors are much lower. This is not the case for 3H, 14C, 35S, iodine and caesium isotopes.
As an example, dose assessment for 137Cs in foods is presented below for the first year after the area
contamination with this nuclide.
For adults: E = 1 000 Bq/kg 550 kg 1.3 10-5 mSv/Bq 0.1 = 0.7 mSv
For infants: E = 1 000 Bq/kg 200 kg 2.1 10-5 mSv/Bq 0.1 = 0.4 mSv

i The import/production factor (IPF) is defined as the ratio of the amount of foodstuffs imported per year from areas
contaminated with radionuclides to the total amount produced and imported annually in the region or country under
consideration.
CXS 193-1995 80

TABLE 2
ASSESSMENT OF EFFECTIVE DOSE FOR INFANTS AND ADULTS
FROM INGESTION OF IMPORTED FOODS IN A YEAR

Guideline level (Bq/kg) Effective dose (mSv)

Radionuclide First year after major

Infant foods Other foods contamination
Infants Adults
238Pu 0.08 0.1
239Pu 0.08 0.1
1 10
240Pu 0.08 0.1
241Am 0.07 0.1
90Sr 0.5 0.2
106Ru 0.2 0.04
129I 100 100 0.4 0.6
131I 0.4 0.1
235U 0.7 0.3
35S* 0.2 0.04
60Co 1 0.2
89Sr 0.7 0.1
103Ru 0.1 0.04
134Cs
1 000 1 000
0.5 1
137Cs 0.4 0.7
144Ce 1 0.3
192Ir 0.3 0.08
3H** 0.002 0.02
14C 1 000 10 000 0.03 0.3
99Tc 0.2 0.4

Notes:
*This represents the value for organically bound sulphur.
**This represents the value for organically bound tritium.
See for “Scientific Justification for the Guideline Levels” (Annex X) and the “Assessment of Human Internal
Exposure when the Guideline Levels are Applied” (Annex XI)
CXS 193-1995 81

ACRYLONITRILE
Reference to JECFA: 28 (1984)
Toxicological guidance value: Provisional acceptance (1984, the use of food contact materials from which acrylonitrile may migrate is provisionally
accepted on condition that the amount of the substance migrating into food is reduced to the lowest level
technologically attainable)
Contaminant definition: acrylonitrile (monomer)
Synonyms: 2-Propenenitrile; vinyl cyanide (VCN); cyanoethylene; abbreviations, AN, CAN.
Related code of practice: Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Commodity/ Guideline level (GL) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
Food 0.02
CXS 193-1995 82

CHLOROPROPANOLS
Reference to JECFA: 41 (1993; for 1,3-dichloro-2-propanol only), 57 (2001), 67 (2006)
Toxicological guidance value: PMTDI 0.002 mg/kg bw (2001, for 3-chloro-1,2-propanediol); maintained in 2006. Establishment of tolerable intake
was considered to be inappropriate for 1,3-dichloro-2-propanol because of the nature of the toxicity (tumorogenic
in various organs in rats and the contaminant can interact with chromosomes and/or DNA).
BMDL 10 cancer, 3.3 mg/kg bw/day (for 1,3-dichloro-2-propanol); MOE, 65 000 (general population), 2 400 (high-
level intake, including young children).
Contaminant definition: 3-MCPD
Synonyms: Two substances are the most important members of this group: 3-monochloropropane-1,2-diol (3-MCPD, also
referred to as 3-monochloro-1,2-propanediol) and 1,3-dichloro-2-propanol (1,3-DCP).
Related code of practice: Code of practice for the reduction of 3-Monochloropropane-1,2-diol (3-MCPD) during the production of acid-
hydrolysed vegetable proteins (Acid-HVPs) and products that contain Acid-HVPs (CXC 64–2008).22
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
Liquid condiments
containing acid-
0.4 The ML does not apply to naturally fermented soy sauce.
hydrolysed vegetable
proteins
CXS 193-1995 83

HYDROCYANIC ACID
Reference to JECFA: 39 (1992), 74 (2011)
Toxicological guidance value: ARfD 0.09 mg/kg bw as cyanide (2011, this cyanide equivalent ARfD applies only to foods containing cyanogenic
glycosides as the main source of cyanide)
PMTDI 0.02 mg/kg bw as cyanide (2011)
Contaminant definition: See explanatory notes in the column “Notes/Remarks”
Synonyms: HCN
Related code of practice: Code of practice for the reduction of Hydrocyanic Acid (HCN) in cassava and cassava products (CXC 73-2013)23
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
The ML is expressed as free hydrocyanic acid.
Gari 2 Whole commodity
The relevant Codex commodity standard is CXS 151-1989.
The ML is expressed as total hydrocyanic acid.
Cassava flour 10
The relevant Codex commodity standard is CXS 176-1989.
CXS 193-1995 84

MELAMINE
Reference to JECFA: FAO/WHO Expert Meeting (2008)
Toxicological guidance value: TDI 0.2 mg/kg bw (2008)
Contaminant definition: Melamine
Commodity/ Maximum level (ML) Portion of the commodity/
Notes/Remarks
Product name mg/kg product to which the ML applies
The ML applies to food other than infant formula.
The ML applies to levels of melamine resulting from its non-intentional
and unavoidable presence in feed and food.
The ML does not apply to feed and food for which it can be proven that
the level of melamine higher than 2.5 mg/kg is the consequence of:
Food (other than
 Authorized use of cyromazine as insecticide. The melamine
infant formulae) and 2.5
feed level shall not exceed the level of cyromazine.
 Migration from food contact materials taking account of any
nationally authorized migration limit.
The ML does not apply to melamine that could be present in the
following feed ingredients/additives: guanidine acetic acid (GAA), urea
and biuret, as a result of normal production processes.
Powdered infant
1
formula
Liquid infant formula 0.15 The ML applies to liquid infant formula as consumed.
CXS 193-1995 85

VINYL CHLORIDE MONOMER

Reference to JECFA: 28 (1984)
Toxicological guidance value: Provisional acceptance (1984, the use of food contact materials from which vinyl chloride may migrate is
provisionally accepted, on condition that the amount of the substance migrating into food is reduced to the lowest
level technologically achievable.
Contaminant definition: Vinylchloride monomer
Synonyms: Monochloroethene, chloroethylene; abbreviation VC or VCM
Related code of practice: Code of practice concerning source directed measures to reduce contamination of foods with chemicals (CXC 49-
2001)16
Portion of the commodity/
Commodity/ Guideline level (GL)
product Notes/Remarks
Product name mg/kg
to which the GL Applies
Food 0.01 The GL in food packaging material is 1.0 mg/kg.
CXS 193-1995 86

NOTES

1 Food and Agriculture Organization (FAO) & World Health Organization (WHO). 2025. Codex Alimentarius Commission
Procedural Manual – Thirtieth edition. Rome.
2 FAO and WHO. 2004 Code of practice on good animal feeding. Codex Alimentarius Code of Practice, No. CXC 54-

2004. Codex Alimentarius Commission. Rome.

3 FAO and WHO. 2007. Working principles for risk analysis for food safety for application by governments. Codex

Alimentarius Guideline, No. CXG 62-2007. Codex Alimentarius Commission. Rome.

4 WHO. 1985. Guidelines for the Study of Dietary Intake of Chemical Contaminants. World Health Organization.

https://iris.who.int/handle/10665/39255
5 FAO and WHO. 1989. Classification of foods and animal feeds. CXA 4-1989. Codex Alimentarius Commission. Rome.
6 FAO and WHO. 2004. Code of practice for the prevention and reduction of aflatoxin contamination in peanuts. Codex

Alimentarius Code of Practice, No. CXC 55-2004. Codex Alimentarius Commission. Rome.
7 FAO and WHO. 2005. Code of practice for the prevention and reduction of aflatoxin contamination in tree nuts. Codex

Alimentarius Code of Practice, No. CXC 59-2005. Codex Alimentarius Commission. Rome.
8 FAO and WHO. 1997. Code of practice for the reduction of aflatoxin B1 in raw materials and supplemental feeding

stuffs for milk producing animals. Codex Alimentarius Code of Practice, No. CXC 45-1997. Codex Alimentarius
Commission. Rome.
9 FAO and WHO. 2008. Code of practice for the prevention and reduction of aflatoxin contamination in dried figs. Codex

Alimentarius Code of Practice, No. CXC 65-2008. Codex Alimentarius Commission. Rome.
10 FAO and WHO. 2017. Code of practice for the prevention and reduction of mycotoxins in spices. Codex Alimentarius

Code of Practice, No. CXC 78-2017. Codex Alimentarius Commission. Rome.

11FAO and WHO. 2003. Code of practice for the prevention and reduction of mycotoxin contamination in cereals. Codex

Alimentarius Code of Practice, No. CXC 51-2003. Codex Alimentarius Commission. Rome.
12 FAO and WHO. 2007. Code of practice for the prevention and reduction of ochratoxin a contamination in wine. Codex

Alimentarius Code of Practice, No. CXC 63-2007. Codex Alimentarius Commission. Rome.
13 FAO and WHO. 2009. Code of practice for the prevention and reduction of ochratoxin a contamination in coffee.

Codex Alimentarius Code of Practice, No. CXC 69-2009. Codex Alimentarius Commission. Rome.
14 FAO and WHO. 2013. Code of practice for the prevention and reduction of ochratoxin a contamination in cocoa.

Codex Alimentarius Code of Practice, No. CXC 72-2013. Codex Alimentarius Commission. Rome.
15 FAO and WHO. 2003. Code of practice for the prevention and reduction of patulin contamination in apple juice and

apple juice ingredients in other beverages. Codex Alimentarius Code of Practice, No. CXC 50-2003. Codex Alimentarius
Commission. Rome.
16 FAO and WHO. 2001. Code of practice concerning source directed measures to reduce contamination of foods with

chemicals. Codex Alimentarius Code of Practice, No. CXC 49-2001. Codex Alimentarius Commission. Rome.
17 FAO and WHO. 2017. Code of practice for the prevention and reduction of arsenic contamination in rice Codex

Alimentarius Code of Practice, No. CXC 77-2017. Codex Alimentarius Commission. Rome.
18 FAO and WHO. 2022. Code of practice for the prevention and reduction of cadmium in cocoa beans. Codex

Alimentarius Code of Practice, No. CXC 81-2022. Codex Alimentarius Commission. Rome.
19 FAO and WHO. 2004. Code of practice for the prevention and reduction of lead contamination in foods. Codex

Alimentarius Code of Practice, No. CXC 56-2004. Codex Alimentarius Commission. Rome.
20 FAO and WHO. 1999. Recommended methods of analysis and sampling. Codex Alimentarius Standard, No. CXS 234-

1999. Codex Alimentarius Commission. Rome.

21 FAO and WHO. 2005. Code of practice for the prevention and reduction of inorganic tin contamination in canned

foods. Codex Alimentarius Code of Practice, No. CXC 60-2005. Codex Alimentarius Commission. Rome.
22 FAO and WHO. 2008. Code of practice for the reduction of 3-Monochloropropane-1,2-diol (3-MCPD) during the

production of acid-hydrolysed vegetable proteins (Acid-HVPs) and products that contain Acid-HVPs. Codex Alimentarius
Code of Practice, No. CXC 64–2008. Codex Alimentarius Commission. Rome.
23
FAO and WHO. 2013. Code of practice for the reduction of Hydrocyanic Acid (HCN) in cassava and cassava products.
Codex Alimentarius Code of Practice, No. CXC 73-2013. Codex Alimentarius Commission. Rome.