Dog Food Heavy Metals: What the 2026 Testing Found

Clean Label Project, a nonprofit, bought 79 top-selling dog foods: 50 dry, 11 air- or freeze-dried, and 18 fresh or frozen. A laboratory produced about 11,376 data points covering heavy metals, pesticides, bisphenols, phthalates, glyphosate, and acrylamide¹. CNN health reporter Sandee LaMotte covered the results on February 12, 2026². The news coverage did not name specific brands². Because the foods were chosen as best sellers, the results describe popular products, not every food on the market.

On the organization's own page, dry food averaged higher than fresh or frozen food for every contaminant listed below¹:

The highest acrylamide reading in a single dry food was 780 parts per billion; the dry-food average was about 48¹. The project also reported that dry food averaged roughly 3 to 13 times the metal levels of human foods it had tested over ten years².

Three cautions apply. First, these are concentrations per serving. A Cornell veterinary nutritionist not involved in the study noted that a dog needs about three cups of fresh or frozen food to match the calories in one cup of kibble, so exposure per calorie from fresh food is higher than the headline ratios imply³. The organization says its per-calorie analysis showed similar differences, so the two views differ¹. Second, the food groups were small, and the authors say they do not know whether country of origin or supply chain explains the differences¹. Third, the study was run and published by an organization that also operates a certification program, and I found no independent replication¹.

In 2018, Cornell researchers measured arsenic, cadmium, lead, and mercury in 51 over-the-counter dry dog foods. Fish-based foods had the most arsenic, cadmium, and mercury, and red-meat foods had the most lead⁴. Dogs took in more of these metals per calorie than average people did, but the authors called chronic toxic exposure “highly unlikely.” They also said an occasional high-lead product deserves caution, especially for puppies⁴. They could not judge toxicity because they measured total metal, not its chemical form⁴.

The FDA Center for Veterinary Medicine has not set guidance or action levels for heavy metals in animal food. It acts case by case, using reference values from the National Research Council and AAFCO⁵. In its 2015-2017 sampling, pet foods contained up to 0.90 ppm lead, below the 10 ppm maximum tolerable level the report lists⁵. That report is from 2019 and used different methods, so it cannot be compared directly with the 2026 numbers. News coverage also cites a 2021 study that reported much higher rates of exceeding limits, but I could not open that study.

Acrylamide forms naturally when starchy plant foods are cooked at high heat. The FDA says high doses caused cancer in lab animals, but those doses were far above what people get from food, and there is no consistent evidence that dietary acrylamide causes cancer in humans⁶. A 2021 Japanese data report found acrylamide averaged 39.6 ng/g in dry dog food versus about 11 ng/g in canned and retort foods. It estimated dogs' daily intake per kilogram at roughly four times that of humans⁷. The abstract does not show how that comparison was calculated, and I found no study showing that acrylamide in food makes dogs ill.

• Do not switch foods abruptly because of a headline. The data do not show that any specific food is unsafe.
• Ask your veterinarian whether your dog's current diet is complete and balanced for its life stage. A fresh diet is not automatically better, and kibble is not automatically unsafe.
• Clean Label Project suggests rotating brands or proteins so that no single source dominates². That is the organization's suggestion, not an established guideline. Ask your vet first, especially for dogs with allergies or health conditions.
• Call your veterinarian about any sudden or unexplained illness in your dog, and mention what you feed.

No one has shown what these levels mean for a dog's long-term health. Coverage of the study notes a lack of long-term feeding studies, and that the safety standards used for pet food were not written specifically for dogs³. A role for food contaminants in canine cancer has been suggested but not demonstrated³. Until independent, peer-reviewed work appears, treat the 2026 results as a signal worth following, not a verdict.