Are dog DNA tests accurate?
The leading tests are accurate at identifying purebred dogs, but not all tests are equal, and no test can predict your dog's personality or diagnose allergies. Here's what independent research shows.
Short answer: in a peer-reviewed 2024 study, Wisdom Panel reported all 12 registered purebred dogs' breeds at 100% and Embark 10 of 12. One company got none right. Mixed-breed results are harder to check, and tests disagree more on small ancestry shares.
The best evidence: six tests, twelve purebred dogs
University of Colorado researchers collected cheek swabs from 12 registered purebred dogs, each a different breed, and sent each sample to six direct-to-consumer companies. To see whether companies lean on the photo owners upload, half the samples went with a photo of a different dog. The results were published in the Journal of the American Veterinary Medical Association in 2024.
| Test | Dogs | Rate |
|---|---|---|
| Wisdom Panel | 12 of 12 | 100% |
| Darwin's Ark | 11 of 11 | 100% |
| Embark | 10 of 12 | 83% |
| DNA My Dog | 8 of 12 | 67% |
| Orivet | 7 of 12 | 58% |
| Accu-Metrics (ViaPet) | 0 of 9 | 0% |
Five of the six tests generally named the registered breed as the main breed, but they often disagreed about secondary breeds. No test showed a statistically significant photo effect, though the authors said the results suggested at least one company was influenced by the photo. DNA My Dog alone reported wolf ancestry in a Bulldog and part Rottweiler in a Beagle. The authors concluded that consumers should approach results with caution.
Human samples: a simple test of quality control
A good lab should be able to say "this sample doesn't work." In a 2024 investigation, a Boston TV reporter's own cheek swab came back from DNA My Dog as Alaskan Malamute, Shar-Pei and Labrador, while Orivet and Wisdom Panel said the sample couldn't give a reliable result. DNA My Dog says it has since moved to a platform that rejects non-dog samples (Global News: DNA My Dog and a human cheek swab (Mar 2024)).
Mixed breeds are harder
Purebreds are the easiest case: there's a registered answer to check against. For mixed-breed dogs there's no such answer, and tests can split small ancestry shares differently. Treat any breed listed at a few percent as a possibility, not a fact.
What DNA tests can't tell you
- Personality. A 2022 study of 18,385 dogs in Science found breed explains about 9% of behavior differences (Scientific American on the 2022 Science study of 18,385 dogs).
- Allergies. At-home hair and saliva allergy kits flagged allergens in stuffed-animal fur and water in a 2018 study (SkeptVet on the Journal of Small Animal Practice study (2018)). DNA tests can't diagnose allergies either.
- A diagnosis. Many disease variants raise risk only a little. Geneticists writing in Nature cited a pug euthanized over a mutation that rarely causes disease (IFLScience on the Nature commentary (2018)). Take any at-risk result to your vet for confirmatory testing.
How to choose an accurate test
- Pick a company with independent evidence: Wisdom Panel and Embark both performed well in the 2024 study.
- Look for clear quality control, such as rejecting samples that aren't dog DNA.
- Choose a health kit that separates at-risk from carrier results and offers professional support.
Our rankings apply these rules: see the best dog DNA tests of 2026, or compare the leaders in Embark vs Wisdom Panel.
Every independent check we track
- 2024: Peer-reviewed: six tests, twelve registered purebreds. University of Colorado researchers published their check in JAVMA. Reporting each dog's registered breed at 100%: Wisdom Panel 12 of 12, Darwin's Ark 11 of 11, Embark 10 of 12, DNA My Dog 8 of 12, Orivet 7 of 12, Accu-Metrics 0 of 9. (JAVMA (2024): six direct-to-consumer tests on 12 registered purebred dogs)
- Mar 2024: A reporter's own cheek swab came back as a malamute mix. WBZ Boston sent a human cheek swab to DNA My Dog, which reported Alaskan Malamute, Shar-Pei and Labrador. Orivet and Wisdom Panel said the sample couldn't give a reliable result. DNA My Dog says it has since moved to a platform that rejects non-dog samples. (Global News: DNA My Dog and a human cheek swab (Mar 2024))
- Jul 2023: Photo test: does the picture change the result?. The same researchers sent half the swabs with a photo of a different dog. DNA My Dog alone reported wolf ancestry in a Bulldog and part Rottweiler in a Beagle, and in one case Accu-Metrics' result looked more like the photo than the DNA. (CU Anschutz: How trustworthy is your dog's DNA test? (Jul 2023))
- Apr 2022: Breed explains about 9% of behavior. A study of 18,385 dogs published in Science found individual dogs vary far more than breed stereotypes suggest. Treat personality-by-breed sections as background, not forecasts. (Scientific American on the 2022 Science study of 18,385 dogs)
- Oct 2018: Hair and saliva allergy tests flagged stuffed-animal fur. Every sample tested positive for allergens, including fake fur and water. No at-home hair or saliva kit can diagnose allergies. (SkeptVet on the Journal of Small Animal Practice study (2018))
- Jul 2018: Geneticists warn: health flags need a vet. A commentary in Nature noted most dog disease-variant tests rest on single studies, and cited a pug euthanized over a mutation that rarely causes disease. (IFLScience on the Nature commentary (2018))