Do Cat DNA Tests Actually Tell You the Breed?

The health panel is the part worth paying for. The breed percentages are not what people think they are.

Why cat breed testing is weak

Dogs were shaped by centuries of intense selection into populations that are genetically very distinct from each other. A dog DNA test has real signal to work with.

Cats were not. Most cat breeds are recent — many formalised in the twentieth century — and several are simply regional populations of ordinary cats given a name. Domestic cats worldwide share an enormous amount of ancestry, and the breeds sit close together on it.

So a test comparing your moggy against a reference panel finds partial matches to several breeds and reports them as percentages. "14% Maine Coon" does not mean a Maine Coon great-grandparent. It means your cat shares markers with the Maine Coon reference group, which most domestic cats in that region also do.

What the tests are genuinely good at

Health markers. This is the real value, and for some cats it is substantial:

• Polycystic kidney disease — widespread in Persians and related breeds
• Hypertrophic cardiomyopathy — breed-specific mutations in Maine Coon and Ragdoll
• Pyruvate kinase deficiency — Abyssinian, Somali, Bengal
• Progressive retinal atrophy — Abyssinian, Bengal
• Glycogen storage disease IV — Norwegian Forest Cat
• Blood type, which matters for transfusions and for breeding

A positive result on any of these is actionable and worth knowing regardless of what the breed section says.

Physical trait genes — coat length, colour, dilution — which are simple and read reliably.

What they cannot do

• Give a meaningful breed breakdown on a mixed cat
• Distinguish breeds that share recent ancestry
• Tell you anything about temperament
• Substitute for a vet's diagnosis. A genetic risk marker is a risk, not a disease.

Reading a result honestly

If it says your cat is 60% domestic shorthair and 40% four different breeds at ten per cent each, that is a cat with no pedigree, described statistically. The 40% is noise around a moggy.

If it says 99% one breed and you bought the cat from a registered breeder with papers, it is confirming what the paperwork already said.

The interesting middle case — a genuine recent pedigree ancestor in a mixed cat — is exactly the case the tests are worst at resolving.

What actually identifies a breed

A pedigree. A breed is a documented ancestry maintained by a registry, and the document is the thing. Everything else — build, coat, head shape, the characters the identification key works on — narrows the possibilities and does not prove them.

The point of that key is not to award your cat a breed. It is to work out what is worth looking at, and to say plainly when the answer is no breed at all, which for most cats it is.

If you do one

• Choose a test that publishes which specific mutations it screens for.
• Take the swab before feeding, and get a proper cheek scrape rather than a quick dab.
• Take the health results to a vet; take the breed results to the pub.

What a vet actually wants from one

If you test, the results worth taking to a consultation are narrow and specific.

Blood type, which matters before any transfusion and before breeding — type B blood in a queen carrying type A kittens causes neonatal isoerythrolysis, which is fatal and preventable.

Polycystic kidney disease status, particularly in a Persian-type cat, because it changes what is monitored and when.

HCM mutation status in a Maine Coon or Ragdoll, which does not replace an echocardiogram but does change how often one is worth doing.

Everything else in the report is background.