Acrylamide Mitigation and Monitoring Readiness (EU 2017/2158)
Apply the right mitigation annex, sample each product type and review measures without delay when a result reaches a benchmark.
In Short
- Benchmarks are not legal maximums.
- Sample known products at least yearly.
- Over benchmark: review measures.
The Food Processing Operations Guide
Safe changeovers. Right labels. Every lot traceable.
- The rules and the plan
- Supplier to filled tray
- Allergen changeovers and pack checks
We ask for your name and email so we can handle this request. The full guide then opens on this site, ready to translate, print or save. Privacy policy. Practical orientation, not legal advice; your quality unit decides which rules apply.
UltimateFree guide
The Food Processing Operations Guide
Safe changeovers. Right labels. Every lot traceable.
Regulation (EU) 2017/2158 has applied since 11 April 2018. It requires producers of listed foods to apply set mitigation measures, and larger operators to sample, analyse and record acrylamide levels against benchmark levels. Benchmark levels are performance indicators, not maximum limits. This guide covers the original regulation's articles and Annexes III and IV. Orientation, not legal advice. Sources checked 4 Oct 2026.
Which foods and which annex
Article 1(2) lists the foods in scope: fried and cut potato products and crisps, potato-dough snacks and crackers, bread, breakfast cereals other than porridge, fine bakery wares such as biscuits, crackers, wafers and gingerbread, roast and instant coffee, coffee substitutes, and baby food and processed cereal-based food for infants. Article 2 then splits operators.
Which annex applies (Article 2)
| Operator | Measures | Own sampling? |
|---|---|---|
| Most producers | Annex I | Yes (Article 4) |
| Retail or local retail supply only | Annex II Part A | No, but show evidence |
| Franchise under central instructions | Annex II Parts A and B | Yes |
- Most producers
- Measures
- Annex I
- Own sampling?
- Yes (Article 4)
- Retail or local retail supply only
- Measures
- Annex II Part A
- Own sampling?
- No, but show evidence
- Franchise under central instructions
- Measures
- Annex II Parts A and B
- Own sampling?
- Yes
More on which foods and which annex
Most producers apply Annex I. Operators that do retail or supply only local retail apply Annex II Part A. Those working under a central operator's brand and instructions, such as franchises, also apply Annex II Part B.
Benchmark levels are indicators, not limits
Article 3 defines benchmark levels as performance indicators used to verify how well the mitigation measures work. Annex IV sets them by category: for example 50 micrograms per kilogram for wheat-based soft bread, 350 for biscuits and wafers, 400 for roast coffee and 40 for baby foods.
Selected benchmark levels (Annex IV)
| Food | µg/kg |
|---|---|
| Wheat-based soft bread | 50 |
| Biscuits and wafers | 350 |
| Crackers (not potato) | 400 |
| Roast coffee | 400 |
| Baby foods (not biscuits) | 40 |
- Wheat-based soft bread
- µg/kg
- 50
- Biscuits and wafers
- µg/kg
- 350
- Crackers (not potato)
- µg/kg
- 400
- Roast coffee
- µg/kg
- 400
- Baby foods (not biscuits)
- µg/kg
- 40
More on benchmark levels are indicators, not limits
Results above a level do not, by this regulation alone, make a product unlawful. Article 2(4) does require operators to review the measures applied and adjust processes and controls, taking account of food safety, production and product conditions. Other rules may apply separately and were not reviewed here.
Your own sampling and analysis programme
Article 4 requires Annex I operators, and Annex II Part B operators, to run a programme of their own sampling and analysis and to record results. Annex III asks for representative samples of each product type, with priority on types that could exceed a benchmark.
More on your own sampling and analysis programme
Products with a known, well-controlled level are sampled at least once a year; higher-risk types more often. Laboratories should take part in proficiency testing, use approved methods and meet the method performance criteria in the annex. Colour or process measures can stand in for analysis only where a statistical correlation is shown.
Records that show the measures were applied
Article 4(2) and (3) require records of the mitigation measures applied, under Annex I or Annex II Parts A and B. Retail and local-supply operators under Article 2(2) are exempt from the sampling rules but must still be able to show they applied Annex II Part A measures.
More on records that show the measures were applied
In practice that means dated evidence such as raw material specifications, recipe and process settings, baking time and temperature, colour checks against a guide and staff instructions. The regulation does not set a retention period in the articles reviewed, so set one in your own procedure.
When a result is above a benchmark
Article 4(5) says that when results are not below a benchmark, the operator reviews its mitigation measures without delay under Article 2(4). A sound review records the result and product type, checks whether the sample was representative, looks at raw materials, recipe, process settings and colour evidence for the period, and decides which changes to make.
Worked Example: Biscuit Line BX-14 Result Review
- Day 0Lab result for BX-14 lot at 410 µg/kg, above 350
- Day 0Technical lead opens review; sample checked as representative
- Day 2Bake profile and colour records for the period compared
- Day 3Team approves lower final-zone temperature; procedure revised
- Next runExtra samples taken to check effect
More on when a result is above a benchmark
People decide the changes and whether any wider action is needed. Article 5 says the Commission reviews benchmark levels every three years, so check for amendments and any later maximum-level rules before relying on the figures here.
Questions
Is a benchmark level a legal maximum?
No. Article 3 calls benchmark levels performance indicators. Exceeding one triggers a review of mitigation measures under Articles 2(4) and 4(5). Check separately whether any later EU rule sets maximum levels for your product.
Do small bakeries have to test for acrylamide?
Operators that do retail or supply only local retail apply Annex II Part A and are exempt from the Article 4 sampling rules, but must be able to show they applied the measures. Franchise-type operators under Part B do sample.
How often should we sample?
Annex III says at least once a year for products with a known and well-controlled level, and more often for products that could exceed the benchmark, based on your own risk assessment per product type.
Can we measure colour instead of acrylamide?
Annex III allows product attributes such as colour or process parameters to replace analysis only where you can show a statistical correlation with acrylamide level.
Practical orientation, not legal advice. Rules differ by jurisdiction; your quality and compliance team decides what applies.
Keep reading
Food CGMP Readiness: Daily Checks and Evidence
BRCGS Food Safety Issue 9 Guide: Culture, HACCP, Traceability and Audits
Cronobacter in Powdered Infant Formula: Part 106 Testing and FDA's Prevention Actions
21 CFR 106 Infant Formula CGMP Guide: Release, Testing and Traceability
Food Allergen Control: US Labeling and Bakery Readiness
Food Fraud and Food Defence: Running VACCP and TACCP Vulnerability Assessments