A few weeks ago, a New York City co-op made headlines when it informed pet-owning residents that they had to produce documentation proving the breeds of their dogs. If the dog was a mix, the percentage of each breed had to be detailed in DNA testing– which prompted cries of “doggie racism,” according to DNAInfo.com. The co-op bans 27 breeds.

Years after dog DNA testing was first introduced, though, it’s finally becoming mainstream. Since Mars Veterinary launched its dog DNA test in 2007, Wisdom Panel, the company — owned by Mars, Incorporated — claims to have sold some 400,000 tests — with the latest consumer version selling for $84.99 a pop. Its other major competitor is DNA My Dog –owned by a Canadian firm– which charges $59.99 per test. Both claim to unlock the mysteries of a dog’s genes to reveal their breeds.

While co-ops and rentals may use the tests if specific breeds are banned, in large part, the sales of DNA tests have been fueled owner curiosity and by animal shelters, which use DNA tests to help place pets into homes. When adopting a pet, prospective owners want to know how big the dogs will get, whether they’re good with kids and if the dogs might be suitable for, say, apartment living. Knowing the breed makeup can shed light on that. Wisdom Panel even makes a shelter test called DogTrax, which gives fast-tracked results since shelter dogs so often have a short amount of time to find a home. Knowing a dog’s breed is also helpful in knowing what health issues for which the dog may be at risk.

But Fortune had a big question: Do these tests actually work? Short of becoming a geneticist and analyzing the DNA yourself, if you have a mixed breed rescue, how can you know whether any of this is legit? I decided to try out the two brands currently available and see if they gave matching results for my dog, Addie.

The author's dog, Addie
My dog, Addie

In 2008, when Hurricane Gustav blew a stray dog I named Addie into my life, she was so thin she resembled a tiny deer. She filled out to look like a small black Lab, but with a fuller tail. I’ve always wondered what breeds she might be.

I used two brands for testing. I used the latest version of Wisdom Panel’s Breed Identification test for owners, Wisdom Panel 3.0, which came out in April, testing for 250 breeds. And 3.0 is the first one to test for coyote and wolf as well.

DNA My Dog is the other consumer test for mixed breed dogs. Located in Toronto, DNA My Dog has sold “lots,” since their 2007 launch according to DNA My Dog’s president Mindy Tenenbaum. DNA My Dog tests for 85 of the most common breeds that Tenenbaum says make up approximately 97% of the common North American mixed breed population. Both tests required taking two samples of cells from the dog’s inner cheek on swabs, which are then air-dried, sealed, and mailed off to the lab.

The results raised more questions, though.

Here’s what the two tests found:

Wisdom Panel: Addie is a Labrador retriever, beagle, bulldog, English springer spaniel cross.

DNA MY Dog: Addie is Labrador retriever (Level 2, meaning 37%-74%), beagle (Level 4, meaning 10%-20%) and English setter (Level 5, the lowest level of 9% or less).

Both sets of results agree on the the two biggest components, as do I: she looks Labrador, and the beagle made sense due to her size and her beagle-like bark. But what of the rest?

 

Addie's family tree, according to DNA testing.
Addie’s family tree, according to DNA testing.

 

So why the different results?

I asked representatives from both companies why the two sets of results differed, and both defended their results. Mindy Tenenbaum from DNA My Dog was confident that Addie’s sample provided all the markers necessary to get a full report, and said if there was any English springer spaniel or bulldog in the mix, they would have reported it. “Although the English setter reported at a low level I see very specific features of the setter in her face,” Tenenbaum said. “She has the ‘look’ of what an English setter often looks like in a mix.” She sent a link to images of English setter/Lab mixes, and some of them did look like my dog.

Katie Lytle, a veterinarian with Wisdom Panel, went over their results for Addie with me, pointing out how her traits were a good visual mix of the four breeds they found. In terms of the discrepancy between brands, she said that all of Addie’s Wisdom Panel results were also breeds that DNA My Dog tests for, but ventured that since Addie has two breeds in her that were raised to hunt waterfowl (Lab and English springer spaniel) “if you think about ancestral groups, English setters were in that class as well, so it’s likely that they [DNA My Dog] were identifying the evidence of the group.”

I also got a bonus tidbit from talking with Dr. Lytle. Looking closer at Addie’s results, she determined that of the three possible types of Labrador Retriever–UK show style, US show style (bred to specifications for show) and US field style (bred for hunting), Addie came from the field style, which made sense given her smaller-than-the-average-Lab stature.

Users of dog DNA tests are often surprised with their results. “Approximately 90% of the time people assume their dog is a border collie, it isn’t,” said Tenenbaum. Many others assume their dog is a Lab mix, when it’s not, she added, since a variety of other mixes can resemble a Labrador. With nothing else to go on but a dog’s looks and behaviors, pet owners who are unsure about their dog’s DNA results may need a follow-up consultation–and both companies offer that option.

Occasionally, skeptics have also tried to test the system by sending DNA samples from the same dog through more than once, under a friend’s name, or they change the gender, said Lytle. Sometimes this can also happen when a dog that’s already been tested by an owner shows up in a shelter that then tests the dog again, or vice versa. Lytle says in these cases, it will pop up as a dog that’s already been analyzed, because each dog’s genetic makeup is unique, like a fingerprint.

Other apparent doubters of the DNA tests’ legitimacy have submitted cat or human cheek cells for analysis, Lytle said. In that case, the sample fails because the lab is testing for dog DNA markers–321 such markers, in the case of Wisdom Panel tests.

I still can only speculate about my dog’s life before we met, but although these test results didn’t line up perfectly, I found that science provided a satisfying amount of information on her heritage. She’s a sporting dog, a hound, and a bully dog–a mutt, just as I suspected. Although according to her Wisdom Panel family tree, all eight great-grandparents were purebreds–that was unexpected.