The Mycotoxin Risk Nobody Talks About in Protein Powders

A pea protein batch we tested last year came back at 18.7 µg/kg for aflatoxin B1 — just under the FDA action level, but nearly five times the EU limit. The brand had no idea. Their contract manufacturer had tested the finished powder for heavy metals and microbial counts, done exactly what was required under 21 CFR Part 111, and signed off on the lot. Mycotoxins weren’t on the panel.

That situation is more common than most people in the supplement industry want to admit.

Protein powders — especially plant-based ones — are made from agricultural commodities: peas, rice, hemp, oats, soy. These crops grow in soil, get harvested and stored under variable conditions, and pass through a supply chain that can span multiple countries before they’re milled into the fine powders consumers scoop every morning. Anywhere along that chain, if temperature and moisture align badly, mold can produce mycotoxins. And unlike bacteria, mycotoxins can’t be eliminated by heating the product. They’re heat-stable, potent, and undetectable by taste or smell at the concentrations that matter.

Why Plant-Based Proteins Carry a Disproportionate Risk

Whey protein — derived from dairy — sits at comparatively low mycotoxin risk. The processing steps involved in cheesemaking and whey concentration, combined with dairy’s properties as a substrate, simply don’t favor the Aspergillus and Fusarium species responsible for most mycotoxin contamination.

For raw material and ingredient-level verification, Ayah Labs specializes in contract testing and supplier qualification.

For EU market entry and European regulatory compliance, Care Europe provides expert consulting from Paris.

Plant-based proteins are a different story. Pea protein is concentrated from field peas (Pisum sativum), which, like all legumes, are harvested at moisture levels that require careful post-harvest drying and controlled storage to prevent mold proliferation. Rice protein is derived from milled rice — a grain with a well-documented aflatoxin and ochratoxin contamination history in high-humidity growing regions across Southeast Asia and South America. Hemp protein is often sourced from Eastern Europe or China, where post-harvest storage infrastructure varies considerably between suppliers.

The mycotoxins most commonly found in these matrices are:

• Aflatoxins (B1, B2, G1, G2) — produced primarily by Aspergillus flavus and A. parasiticus
• Ochratoxin A (OTA) — produced by Aspergillus ochraceus and Penicillium verrucosum
• Fumonisins (B1, B2) — associated with corn and rice, produced by Fusarium species
• Deoxynivalenol (DON) — a Fusarium toxin prevalent in oat- and wheat-derived proteins
• Zearalenone (ZEN) — an estrogenic mycotoxin found in grains, increasingly flagged in rice protein concentrate

A 2019 survey published in Food Additives & Contaminants found at least one mycotoxin above the limit of quantification in roughly 40% of commercially available plant-based protein supplements tested. That’s not a fringe concern — that’s a statistically meaningful slice of the market, and the data aligns with what we see coming through our own intake queue.

Aflatoxin vs. Ochratoxin A: Different Toxicology, Same Testing Gap

These two toxins dominate the conversation for good reason.

Aflatoxin B1 is classified as a Group 1 human carcinogen by the International Agency for Research on Cancer (IARC) — the same category as tobacco smoke and benzene. It’s primarily hepatotoxic, with chronic low-level exposure strongly associated with hepatocellular carcinoma. The FDA action level for total aflatoxins in human food is 20 µg/kg (20 ppb). The EU takes a considerably stricter position: 4 µg/kg for aflatoxin B1 alone, and 10 µg/kg for total aflatoxins in foods for direct human consumption.

That gap creates a real compliance problem. A product that clears the FDA action level can simultaneously be in violation of EU regulations. And regulators on both sides of the Atlantic pull market samples.

Ochratoxin A operates differently. It’s nephrotoxic — the kidneys bear the primary burden — and classified as a Group 2B possible human carcinogen. More practically, OTA has a biological half-life in humans of approximately 35 days, meaning repeated low-level exposure accumulates in tissue over time rather than clearing quickly between doses. The EU sets limits of 3 µg/kg in raw cereals and 10 µg/kg in dried vine fruits. There’s no specific FDA action level for OTA in dietary supplements, which creates an ambiguous compliance environment that we see reflected in the wide variation of what brands choose — or don’t choose — to test for.

Here’s what often surprises clients: a pea protein batch can present with acceptable moisture content, clean microbial plate counts, and no visible discoloration, and still carry meaningful ochratoxin A levels. The mold that produced the toxin may have been killed during processing. The toxin it left behind was not.

What a Complete Mycotoxin Panel Actually Looks Like

Not all mycotoxin testing is equivalent, and a panel that only screens for total aflatoxins by lateral flow immunoassay is inadequate for a plant-based protein brand making any kind of quality or clean-label claim.

A complete panel for protein powders should quantify:

1. Aflatoxins B1, B2, G1, and G2 — reported individually, not just as a total
2. Ochratoxin A
3. Fumonisin B1 and B2 — particularly relevant for rice and corn-derived proteins
4. Deoxynivalenol (DON) — relevant for oat-, wheat-, or multigrain-based formulas
5. Zearalenone — worth including for any rice or grain-derived ingredient

The analytical gold standard for this full profile is LC-MS/MS (liquid chromatography–tandem mass spectrometry), which reliably quantifies these compounds at 0.1–1.0 µg/kg — well below regulatory action levels and well below the typical quantitation range of immunoassay screens. ISO 16050 (determination of aflatoxin B1 in cereals) and AOAC Official Methods for mycotoxins in food matrices provide the validated frameworks most labs use as a starting point, with method modifications applied for the protein-dense matrices that characterize these ingredients.

At Qalitex, we run full multi-mycotoxin panels by LC-MS/MS as part of our dietary supplement raw material testing protocol. One thing we consistently see is matrix interference from high-protein samples suppressing signals in simpler ELISA-based screens — the false-negative risk isn’t trivial, and it’s one reason we don’t rely on immunoassay as a sole confirmation tool for protein powders.

Screening by ELISA can be cost-effective for high-volume incoming material checks, but it should never be the final word. Confirmatory LC-MS/MS — especially for new supplier qualifications or ingredients from regions with variable weather in a given crop year — is worth the additional 24–48 hours of turnaround.

Reading Your Results: What the Numbers Mean in Practice

Say your rice protein concentrate comes back at 3.8 µg/kg for aflatoxin B1 and 7.2 µg/kg for fumonisin B1. What do you do with that?

For the aflatoxin B1 result: 3.8 µg/kg sits below FDA’s 20 ppb action level, but it exceeds the EU limit of 4 µg/kg. If your product is U.S.-only and you have no international distribution, you’re technically within domestic compliance. But if your product is available on Amazon’s international storefronts, sold through a UK distributor, or exported anywhere in the EU — that batch is a failure, not a pass.

For the fumonisin B1 result: there’s no specific EU or FDA action level for fumonisins in rice protein specifically. FDA’s 2001 guidance recommends ≤2,000 µg/kg for fumonisin-containing human food products from corn. At 7.2 µg/kg, this result isn’t a regulatory concern — but it does indicate measurable field fungal pressure in the source grain. That’s worth a conversation with your supplier about post-harvest drying conditions and storage duration before the next purchase order.

Context and trend data matter as much as the individual result. One slightly elevated lot is a data point. Three consecutive lots from the same supplier trending upward is a supplier qualification problem.

The Sourcing Variable That Resets Without Warning

One pattern we see repeatedly: a brand switches to a new ingredient supplier to reduce costs, doesn’t update its incoming raw material testing protocol to cover the new origin region, and discovers six months later that the new pea protein consistently tests at the higher end for OTA. The contamination was present from batch one. Nobody caught it because the testing specification hadn’t been revisited.

This isn’t unusual negligence — it’s the practical reality that mycotoxin risk is dynamic. It shifts with crop year, growing region, weather patterns during the growing season, and post-harvest handling at individual facilities. A supplier who delivered clean material for two consecutive years can have a problematic harvest in year three. The 2022 European drought drove significantly elevated aflatoxin levels in Southern European maize and cereal crops, for instance, with Aspergillus infection rates rising sharply in heat-stressed grain. Brands sourcing European oat protein that year who hadn’t tested the new-crop material against a mycotoxin specification found out late.

Under 21 CFR Part 111, dietary supplement manufacturers must establish and test against raw material specifications before use. But the regulation doesn’t specify which contaminants those specifications must address. That gap is filled — or not filled — by each brand individually. The ones who build a multi-mycotoxin specification into every raw material certificate of analysis requirement are the ones who intercept these lots before they’re processed and packaged.

What You Should Do Before the Next Purchase Order

The global protein supplement market reached approximately $21 billion in 2024 and is growing at roughly 8% annually, driven largely by plant-based product demand. That consumer base is increasingly informed, proactively testing products independently, and not quiet when results are concerning. A mycotoxin failure isn’t just a regulatory event — it’s a brand event, and it’s one of the few food safety risks that a brand can substantially control through the right incoming material testing program.

If you’re formulating, co-manufacturing, or sourcing ingredients for a protein powder, here’s where to start: require a full multi-mycotoxin panel by LC-MS/MS on every new supplier qualification — not ELISA alone, and not just aflatoxins. Set specifications that reflect both FDA action levels and EU limits if there’s any plausible international distribution channel. Revisit those specifications every time you change suppliers, shift your ingredient to a new country of origin, or hear about unusual weather in a key growing region during the prior growing season.

The mold that made the toxin is long gone by the time the powder reaches you. The problem it left behind is yours to find before your customer does.