Preservatives in pet food: BHA, BHT and ethoxyquin, a safety guide
Antioxidants stop the fat in a kibble from going rancid, and rancidity is itself undesirable. The controversy is not whether preservatives are needed but which ones, and at what margin of safety. Three names dominate the debate: BHA and BHT, two synthetic antioxidants permitted on both sides of the Atlantic, and ethoxyquin, which the European Union has refused to authorise while the United States still permits it under limits. The honest reading separates the intrinsic hazard of a substance from its risk at the doses actually used.
Last updated :General documentary information. For an individual animal, a veterinarian's advice takes precedence over any online content.
This guide sets the agency positions side by side, keeps EU status distinct from US status, dates each regulatory step, and gives the level of evidence throughout. It also explains the natural alternatives and how a buyer can infer the preservation system when the label stays vague.
On this page (Preservatives food)
- Are the BHA and BHT in pet food carcinogenic?
- What do EFSA, IARC and the NTP actually conclude?
- Is ethoxyquin still allowed, and where?
- Why did the EU refuse ethoxyquin?
- Are natural preservatives safer or just shorter-lived?
- How can a buyer read the preservation system on a label?
- Comparison: the main pet food antioxidants at a glance
- The verdict: dose and origin decide, not the chemical family
Are the BHA and BHT in pet food carcinogenic?
Answer capsule. The debate is open, not settled. The US National Toxicology Program lists BHA as "reasonably anticipated to be a human carcinogen", a status held since 1991 and based on forestomach tumours seen in rodents at high doses (NTP, Report on Carcinogens). At permitted use levels no causal link to cancer is established for BHA or BHT, and BHT does not carry the same classification.
The crucial nuance is anatomical and dosimetric. The rodent forestomach, where those tumours arose, has no direct equivalent in the dog, the cat or humans, which limits any mechanical transposition. And toxicology always separates the experimental dose, very high, from the use dose, low and capped by regulation. BHA and BHT remain authorised antioxidants in EU animal feed (E320 and E321) and hold GRAS status in the United States (EFSA; FDA).
The evidence for harm at real exposure is low to uncertain, because the carcinogenicity data come from doses far above anything an animal would receive. That justifies caution about the safety margin without asserting a proven hazard at legal doses.
What do EFSA, IARC and the NTP actually conclude?
Answer capsule. They converge on caution and diverge in method. EFSA reasons by acceptable daily intake and has confirmed reference values for BHA and BHT, judging their controlled use acceptable. IARC classes BHA in group 2B, "possibly carcinogenic", and BHT in group 3, "unclassifiable". The NTP lists BHA as "reasonably anticipated". No agency concludes a proven hazard at dietary doses (EFSA; IARC; NTP).
Reading these labels correctly matters. Group 2B is a category of limited evidence; it has at various times included items as ordinary as whole-leaf aloe vera extract, which shows that 2B signals uncertainty rather than a demonstrated risk. EFSA's approach is different in kind: it sets a threshold below which exposure carries no appreciable risk, then keeps real use far below the doses tested in studies.
So the agencies are not contradicting each other; they are answering different questions. IARC asks how strong the evidence of hazard is, EFSA asks what intake is acceptable. The evidence differs between bodies but points the same way: caution, no established hazard for the dog or cat at dietary doses.
Is ethoxyquin still allowed, and where?
Answer capsule. Not in the European Union. Ethoxyquin (E324) is no longer authorised as an EU feed additive: Regulation (EU) 2022/1375, applicable since August 2022, definitively refused its authorisation (EUR-Lex). In the United States the FDA still permits its use within maximum limits. The product's regulatory origin is therefore decisive.
The file has two distinct dates that are easy to conflate. The EU first suspended ethoxyquin in 2017 under Regulation 2017/962, for lack of sufficient data, then turned that suspension into a definitive refusal in 2022. A suspension is reversible if reassuring data arrive; a refusal closes the route unless the file is reopened. Between 2017 and 2022 the additive was already unusable in the Union, so the 2022 decision mostly confirmed a five-year status quo.
The consequence for a buyer is concrete: the same additive can be banned in an EU-made food and legal in a US-made food of comparable recipe. The evidence is regulatory and dated, which is why citing 2017 for the suspension and 2022 for the refusal removes all ambiguity.
Why did the EU refuse ethoxyquin?
Answer capsule. Because safety could not be demonstrated. In its opinion of 3 March 2022, EFSA could not rule out the genotoxicity of p-phenetidine, an impurity remaining after manufacture, and flagged uncertainties for long-lived animals and possible aquatic contamination (EFSA, 2022). That unresolved uncertainty, not a proven hazard, grounded the refusal on precautionary grounds.
The European precautionary principle does not require proof of harm; it requires that a substance not be authorised when its safety is not established (EUR-Lex). A telling detail: the refusal applies even though the additive was judged safe for some farmed species, which shows it was residual uncertainty about an impurity that decided the matter. EFSA assesses, the Commission decides, and the inconclusive opinion fed directly into Regulation (EU) 2022/1375.
A practical wrinkle explains decades of owner concern. Ethoxyquin long stabilised fish meal at the supplier stage, and labelling law requires declaring only additives added to the finished product, not those carried in a raw material bought already stabilised (Regulation (EC) 767/2009). Since the 2022 refusal, an EU-compliant food must no longer contain it, including through its raw materials, so oversight has shifted upstream onto suppliers. The evidence is toxicological and regulatory in combination.
Are natural preservatives safer or just shorter-lived?
Answer capsule. Mostly shorter-lived. Mixed tocopherols (vitamin E), rosemary extract and ascorbic acid (vitamin C) genuinely protect fats from oxidation, but their action fades faster and tolerates extrusion heat and oxygen less well than synthetic antioxidants (FEDIAF, 2024; Tufts Petfoodology). The practical result is a closer best-before date, with no demonstrated safety gain either way.
Tocopherols are worth a clarification because they confuse many readers. They are the chemical family of vitamin E and play a dual role: an antioxidant that protects the fat in the bag and an essential nutrient in the body. The wording "preserved with mixed tocopherols (source of vitamin E)" therefore signals a natural antioxidant system, coded E306 to E309 in the EU (EFSA). The quantity added as a preservative is low and does not, on its own, cover the nutritional requirement.
The trade-off is between prolonged stability and a preference for plant-origin additives, not between danger and safety. A bag opened then poorly resealed re-admits oxygen and restarts oxidation that natural antioxidants struggle to contain once activated, so post-opening storage matters as much as the antioxidant chosen. The evidence on lower persistence is strong and shows up directly in printed shelf lives.
How can a buyer read the preservation system on a label?
Answer capsule. By reading the additives section and the best-before date together. Synthetic antioxidants appear as "BHA", "BHT" or the codes E320 and E321; natural ones as "tocopherols" or "rosemary extract". A shorter shelf life, often 12 to 18 months, frequently points to a natural system (Tufts Petfoodology; Regulation (EC) 767/2009).
For completeness, note that propyl gallate (E310), BHA (E320) and BHT (E321) are all authorised EU feed antioxidants, with propyl gallate re-authorised for all species by Regulation (EU) 2025/281 under a maximum content. That three chemically related antioxidants can sit beside ethoxyquin, which was refused, is the clearest illustration of the rule: authorisation is judged compound by compound and dose by dose, never by chemical family (EFSA, 2020; EUR-Lex).
One blind spot remains. An antioxidant may be present in a raw material such as fish meal without appearing on the finished label if it was added upstream, the same mechanism that long concealed ethoxyquin. When the label stays vague, asking the manufacturer about the preservation system and the origin of any fish meal is the most reliable route. The evidence here is practical rather than toxicological.
Comparison: the main pet food antioxidants at a glance
The table summarises status and evidence for the antioxidants most discussed, keeping EU and US frameworks distinct. The headline point is that "synthetic" and "natural" do not map onto "risky" and "safe"; status is assigned substance by substance.
| Antioxidant | EU status (feed) | US status | Hazard classification | Causality at use doses |
|---|---|---|---|---|
| BHA | Authorised (E320) | GRAS, low doses | NTP "reasonably anticipated"; IARC 2B | Not established |
| BHT | Authorised (E321) | Permitted, low doses | IARC group 3 (unclassifiable) | Not established |
| Propyl gallate | Authorised (E310, Reg. (EU) 2025/281) | Permitted | Assessed acceptable (EFSA, 2020) | Not established |
| Ethoxyquin | Refused (Reg. (EU) 2022/1375) | Permitted with limits | Impurity genotoxicity not excluded | Safety inconclusive (EFSA) |
| Mixed tocopherols | Authorised (E306 to E309) | Permitted | No comparable concern | Not in question |
The verdict: dose and origin decide, not the chemical family
The evidence supports a measured position rather than alarm or complacency. BHA and BHT carry hazard flags drawn from high-dose animal studies, but no agency demonstrates a causal cancer link at the low, capped doses used in feed (NTP; EFSA; IARC; FDA). Ethoxyquin is the genuinely settled case in the EU, refused on precautionary grounds because an impurity's genotoxicity could not be ruled out (EFSA, 2022; Regulation (EU) 2022/1375), yet it remains legal in the United States.
The practical recommendation is to let regulatory origin and the label do the work. For an EU buyer, choosing a food made under EU rules already excludes ethoxyquin, including via raw materials. If you prefer to avoid synthetic antioxidants, look for "tocopherols" or "rosemary extract" and accept a closer best-before date, then store the bag sealed and dry to protect the fat. Above all, do not treat a chemical family as uniformly risky or uniformly safe: the status of each antioxidant rests on its own toxicological file and a dated text, which is exactly how this guide has read them.
---
Keep reading
Related questions: Are the BHA and BHT in pet food carcinogenic? | Is ethoxyquin still permitted in pet food in Europe? | What natural preservatives exist as alternatives to BHA and BHT?
Glossary: Ethoxyquin (E324) | Tocopherols (vitamin E)
Hub: Controversial ingredients: myths versus evidence
Sources: NTP, Report on Carcinogens (BHA); EFSA, re-evaluations of BHA (E320), BHT (E321), propyl gallate (2020) and ethoxyquin (opinion of 3 March 2022); IARC (CIRC) Monographs; FDA, GRAS status and ethoxyquin; Regulation (EU) 2022/1375, Regulation 2017/962 and Regulation (EU) 2025/281 (EUR-Lex); FEDIAF, EU Legislation; Tufts Petfoodology.