Mycotoxins in Herbal Ingredients: EU Limits Under Regulation 2023/915 and What Your Export Markets Demand

In 2024, the EU’s Rapid Alert System for Food and Feed (RASFF) logged over 340 notifications involving mycotoxin contamination — the single largest category of chemical-hazard alerts for the third consecutive year. A disproportionate share of those notifications involved dried herbs, spices, and botanical raw materials. And yet, when European contract manufacturers and ingredient suppliers bring us their quality dossiers, the mycotoxin testing section is almost always the thinnest part of the file.

That gap is a commercial problem, not just a compliance one. US importers, Canadian NHP licensees, and major European retail chains are all tightening their incoming raw material requirements. If your certificate of analysis doesn’t show the right mycotoxins, tested by the right method, from an accredited laboratory, you’re going to start losing purchase orders — quietly, and without much explanation.

What EU Regulation 2023/915 Actually Requires for Herbal Ingredients

Commission Regulation (EU) 2023/915 of 25 April 2023 replaced Regulation (EC) No 1881/2006 as the consolidated framework for maximum levels (MLs) of contaminants in food. For herbs, spices, and botanical ingredients placed on the EU market, the relevant limits are:

• Aflatoxin B1: 2 µg/kg
• Sum of aflatoxins (B1 + B2 + G1 + G2): 4 µg/kg
• Ochratoxin A: 15 µg/kg in dried spices; 10 µg/kg in dried herbs used as food
• Patulin: applies to apple-derived botanicals and some fruit-based preparations

These aren’t aspirational thresholds — they’re legally enforceable maximum levels that apply to any food business operator placing these materials on the EU market. Exceed them, and you face product withdrawal, a RASFF notification, and rapid removal from customer supplier lists. The burden of proof rests with the manufacturer or importer: you must demonstrate compliance before the product ships, not after a border control sample fails.

Sampling and analysis methods are governed separately under Commission Regulation (EC) No 401/2006 and its subsequent amendments. That regulation specifies minimum sampling plans — stratified random sampling by lot weight — and performance criteria for analytical methods. It’s worth reading alongside 2023/915, because results are only defensible if the sampling was performed correctly. A perfect LC-MS/MS result on a non-representative subsample is still an audit finding waiting to happen.

Why Standard European Food Labs Often Miss Mycotoxins in Botanical Matrices

Here’s the part that catches manufacturers off guard: testing for mycotoxins in wheat or maize is genuinely different from testing in ashwagandha root powder, echinacea, or licorice extract. Botanical matrices are chemically complex. Polyphenols, tannins, essential oils, and plant pigments interfere with the antibody-binding mechanisms that immunoassay methods — ELISA, lateral flow strips — rely on. Studies comparing immunoassay to LC-MS/MS in herbal matrices have documented false-negative rates of 15–30%, depending on the specific matrix and mycotoxin combination.

LC-MS/MS (liquid chromatography coupled with tandem mass spectrometry) is the reference method under EU 401/2006 for good reason. It achieves detection limits of 0.1–0.5 µg/kg — well below the MLs in 2023/915 — and handles matrix interference far more reliably than immunoassay screens. The challenge is that relatively few European laboratories hold ISO 17025 accreditation with validated LC-MS/MS methods specifically scoped to botanical and herbal matrices. Many labs have accreditation for cereals and nuts, and then apply those cereal-validated methods to turmeric powder without flagging that the validation was never performed on that matrix.

We’ve seen this cause real commercial damage. A batch of ginger root powder passes ELISA screening at a European food lab, gets released, ships to a US supplement brand, and fails incoming LC-MS/MS testing at aflatoxin B1 = 3.8 µg/kg. The EU test missed it. The client loses the order, absorbs destruction costs, and spends months investigating a supply chain that looked clean on paper. That scenario is not a hypothetical — it’s a pattern.

What US and Canadian Markets Add to the Equation

EU compliance is necessary for European market access, but it doesn’t get your product across the Atlantic on its own.

In the United States, the FDA has no single published numeric mycotoxin limit for dietary supplement ingredients under 21 CFR Part 111. Manufacturers are required to establish written specifications for every raw material, backed by validated test results. In practice, the industry has converged on:

• Total aflatoxins ≤ 20 µg/kg (following the FDA’s action level for food, adopted by convention for supplements)
• Ochratoxin A ≤ 10 µg/kg (the European ML used as a de facto benchmark, since FDA has no formal limit)
• Individual brands often apply tighter internal limits — 5 µg/kg total aflatoxins isn’t unusual for clean-label positions

USP Chapter <561> (Articles of Botanical Origin) and American Herbal Pharmacopoeia monographs provide additional testing guidance that sophisticated US buyers reference when qualifying ingredient suppliers. If your COA doesn’t include a mycotoxin panel run by an ISO 17025-accredited lab using LC-MS/MS, you’re not getting onto most approved supplier lists — particularly for brands selling through Amazon, which can pull product listings within hours of a failed third-party audit.

In Canada, the Natural Health Products Regulations (SOR/2003-196) require that raw materials used in licensed NHPs meet the contaminant limits prescribed under the Food and Drug Regulations. Health Canada’s guidance on Good Manufacturing Practices for NHPs explicitly specifies incoming raw material testing — including mycotoxin panels — as a required element of the quality management system. Canadian NHP licensees are legally responsible for their supplier qualification program, which means they will push that testing burden directly onto you as the ingredient supplier.

The combined effect is this: a well-designed, comprehensive mycotoxin panel run by an accredited lab on each incoming lot is no longer optional for manufacturers with multi-market ambitions. It’s the cost of doing business.

Building a Mycotoxin Testing Protocol That Holds Up Across Markets

Based on our work qualifying European ingredient suppliers for North American export and EU retail customers, a defensible protocol has five core elements.

Define your panel before you source. At minimum for botanical ingredients: aflatoxins B1, B2, G1, and G2; ochratoxin A; fumonisins B1 and B2 (particularly relevant for grain-derived botanicals); deoxynivalenol (DON); and zearalenone. High-risk materials — turmeric, ginger, black pepper, fenugreek, and licorice root — have well-documented mycotoxin susceptibility patterns and warrant fuller panels on every incoming lot.

Check your lab’s accreditation scope, not just its certificate. ISO 17025 accreditation is the starting point. But the scope annex matters more. Ask for the scope document and read the listed test matrices. If the lab’s scope specifies “cereals and cereal products” and you’re submitting fennel seed or valerian root, their results are not defensible under Regulation 401/2006. This is a detail that slips past quality managers more often than it should.

Specify the method in your supplier quality agreement. LC-MS/MS for confirmatory testing, written into the contract. Lateral flow or ELISA as a preliminary internal screen is acceptable, but any COA released to customers must reference LC-MS/MS confirmation. Put it in writing — in your supplier agreements, your incoming inspection SOPs, and your product specification sheets.

Risk-stratify your sampling frequency. Botanical materials from South and Southeast Asia, sub-Saharan Africa, and parts of South America carry systematically higher contamination risk due to climate, post-harvest storage conditions, and handling infrastructure. For those origins, every-lot testing is prudent. For established suppliers from lower-risk origins with a clean testing history, skip-lot protocols — typically every third lot after five consecutive passing results — can be justified, provided you document the risk rationale in your quality risk assessment or HACCP file.

Align your acceptance limits to your most demanding market. For total aflatoxins, the EU limit of 4 µg/kg is significantly stricter than the US convention of 20 µg/kg. If you’re shipping to both markets from the same production batch, use 4 µg/kg as your release criterion. For ochratoxin A in herbs, 10 µg/kg covers both EU and Canadian requirements. Apply the tighter limit at the point of raw material acceptance.

The Practical Reality for European Botanical Manufacturers

Most EU manufacturers of herbal ingredients and finished botanical products are underinvesting in mycotoxin testing relative to what Regulation 2023/915 requires — let alone what North American export markets demand. The gap isn’t deliberate; it comes from limited lab capacity for botanical matrices in Europe, combined with a regulatory framework that places the testing burden on the food business operator without spelling out exactly what that means in method, matrix, or frequency terms.

Many manufacturers assume their ingredient suppliers are handling it upstream. Some are. Many aren’t — and without a documented supplier qualification program that includes verified COAs with accredited mycotoxin results, that assumption won’t hold up in an inspection.

The audit pressure is increasing. US brands are tightening supplier COA requirements following a documented uptick in FDA import refusals for herbal ingredients with mycotoxin exceedances in 2023–2024. Large Canadian NHP licensees are now requesting method validation reports alongside numeric results during supplier qualification visits — not just the numbers, but proof that the lab’s method was validated for your specific matrix.

A well-designed protocol isn’t a complex undertaking — it’s a specific one. Get the panel right, get the method right, confirm the accreditation scope, and document your risk rationale. If you want help mapping your current testing program against the requirements of EU 2023/915, FDA 21 CFR Part 111, or Health Canada’s NHP GMP guidance — or if you need to identify an ISO 17025-accredited lab with validated methods for your specific botanical matrices — that’s exactly the kind of gap analysis we do.

---

Written by Nour Abochama, Quality & Regulatory Advisor, Care Europe | VP Operations, Qalitex. Learn more about our team

Talk to our team about EU market entry and export compliance for herbal ingredients. Contact us

Related from our network

• ISO 17025 Mycotoxin Testing for Botanical Ingredients in the US Market — Qalitex Laboratories provides validated LC-MS/MS mycotoxin panels for herbal and botanical matrices, supporting US supplement brands and European exporters entering the American market.
• Health Canada NHP Raw Material Testing and Supplier Qualification — Androxa supports Canadian NHP licensees with incoming raw material testing — including mycotoxin panels required under Health Canada’s NHP GMP guidance — for European ingredient suppliers qualifying into the Canadian market.