Which Dog Breeds Are Most Likely to Be Returned to Shelters (And Why)

The only large-scale peer-reviewed study to actually measure this found toy breeds were 65% less likely to be returned than herding breeds, and terriers were 35% less likely — but the bigger factor wasn’t breed at all. An adult dog between 2 and 8 years old was over three times more likely to be returned than a puppy, a gap far larger than any difference between breed groups.

What the research actually shows, in short:

• Toy breeds (OR 0.35) and terriers (OR 0.65) were significantly less likely to be returned than herding breeds; hound, sporting, working, non-sporting, and pit bull-type dogs showed no statistically significant difference from herding breeds on a first return.
• Age predicted return far more strongly than breed: dogs 2–8 years old had 3.4x the odds of return of a puppy, and even senior dogs (8+) had more than double a puppy’s odds.
• Pit bull-type dogs weren’t more likely to be returned the first time — but they were significantly overrepresented among dogs returned multiple times, and had 2.6x the odds of being euthanized after a return.
• A separate, earlier study found dogs labeled “pit bull” in a shelter took roughly three times longer to be adopted than visually similar dogs without that label, even though adopters rated their photos equally attractive once the label was removed — suggesting part of the disparity is about perception, not just behavior.

What Does the Only Large-Scale Study on This Actually Measure?

Most of what circulates online about “worst breeds for shelter returns” is guesswork repeated from other guesswork. The actual data comes from one study: Lauren Powell, Chelsea Reinhard, and colleagues at the University of Pennsylvania published “Characterizing unsuccessful animal adoptions” in Scientific Reports in 2021, based on records from Charleston Animal Society in North Charleston, South Carolina, covering every adoption and return between January 2015 and December 2019. Of 9,996 dogs adopted during that period, 1,628 came back — a 16.3% return rate. That’s a real, substantial dataset, but it’s still one shelter in one region, so treat the numbers below as the best available evidence, not a national guarantee for every dog of a given breed.

Which Breed Groups Are Actually Least and Most Likely to Be Returned?

Using herding breeds as the comparison point, the study found toy breeds were 65% less likely to be returned (odds ratio 0.35) and terrier breeds were 35% less likely (odds ratio 0.65) — both statistically significant results. Every other breed group tested — hound, sporting, working, non-sporting, and pit bull-type — showed no statistically significant difference in first-return likelihood compared with herding breeds. In plain terms: the data supports saying toy and terrier breeds are a comparatively safer bet against a first return, but it does not support ranking the remaining breed groups against each other the way a lot of “worst dog breeds for shelters” lists imply. Herding breeds showing up as the reference group with the highest return odds likely reflects a mismatch between their exercise and mental-stimulation needs and the average adopter’s lifestyle, rather than herding dogs being inherently harder to live with.

Does Breed Actually Matter Less Than Age?

This is the part most coverage of this topic skips entirely. The same study found that a dog’s age at adoption predicted return far more strongly than its breed group did. Compared with puppies, dogs between 2 and 8 years old had 3.4 times the odds of being returned, young adults between 6 months and 2 years had 2.9 times the odds, and even senior dogs over 8 had more than double the odds. None of the breed-group effects came close to that size. If you’re trying to reduce the odds of a return — for the dog’s sake as much as your own — the data says matching the dog’s age and energy level to your actual routine matters more than which breed group it falls into.

What’s Really Going on With Pit Bull-Type Dogs?

This is the finding that most benefits from being stated precisely, because the simplified version gets repeated inaccurately online. Pit bull-type dogs were not statistically more likely to be returned the first time in the Charleston data. What the study did find is that pit bull-type dogs were significantly overrepresented among dogs returned to the shelter more than once, and that they had 2.6 times the odds of being euthanized following a return compared with other breeds. The gap isn’t in whether these dogs get a first chance — it’s in what happens after a setback.

Part of that gap traces to perception rather than behavior. A 2016 study by Lisa Gunter, Rebecca Barber, and Clive Wynne, published in PLOS ONE, found that dogs labeled “pit bull” on their kennel cards took roughly three times longer to be adopted than visually similar dogs given a different label, and that removing breed labels entirely was associated with a 12% reduction in euthanasia for pit bull-type dogs. When the same researchers showed adopters photos of the dogs without any breed label attached, there was no difference in how attractive the dogs were rated — meaning the slower adoptions and harsher outcomes tracked the label itself, not anything visibly different about the dogs. Combined with the Charleston findings, the honest picture is that pit bull-type dogs face a harder road back into a home after a return, driven by a mix of real behavioral screening and a documented labeling bias that affects them specifically.

Why Do Dogs Actually Get Returned, Regardless of Breed?

Breed group was a modest factor; the reasons owners actually gave for returning dogs were dominated by two categories. Behavioral issues accounted for 36.1% of dog returns in the Charleston data, and incompatibility with existing pets in the home accounted for another 18.3% — together covering more than half of all returns, across every breed group. That second category lines up closely with what shows up in the clinical research on household conflict between resident dogs and new additions: a newer or younger dog is disproportionately the one that triggers friction, which can end in a return regardless of which breed group it belongs to.

Head-to-Head: What the Data Actually Shows

How to Use This Data If You’re Choosing a Breed

1. Weigh age and lifestyle fit at least as heavily as breed group — the data shows it’s a bigger predictor of whether an adoption sticks.
2. Don’t treat “herding breed” or “working breed” as a red flag in itself; the elevated return odds for those groups most likely reflect an exercise and stimulation mismatch, which is solvable with an honest look at your actual daily routine before adopting, not a reason to rule the group out.
3. If you’re adding a second dog rather than a first, plan for the introduction period deliberately — incompatibility with an existing pet is one of the two leading return reasons regardless of breed, and most of that friction is manageable with the right setup rather than inevitable.
4. If you’re drawn to a pit bull-type dog, know that the data-backed risk isn’t in bringing the dog home — it’s that a setback or a return can be treated more harshly by the system than the same setback would be for another breed, which argues for lining up a trainer or behaviorist early rather than waiting for a crisis.
5. Treat any specific dog’s behavior evaluation and your own routine as more informative than its breed label — the research above shows the label itself can distort outcomes independent of the dog’s actual behavior.

If you’re weighing a breed specifically because you work full-time, it’s also worth cross-checking the practical lifestyle fit covered in this guide to dog breeds for 9-5 schedules against the age guidance above — a lower-energy adult dog can often fit a working schedule better than a high-energy puppy of a “low maintenance” breed.

Frequently Asked Questions

What breed is most likely to be returned to a shelter?

The only large-scale study on this found herding breeds had the highest return odds among the groups compared, though the difference was only statistically significant against toy breeds and terriers specifically — not against every other group.

Are pit bulls more likely to be given up than other dogs?

Not on a first return — the data found no statistically significant difference there. Pit bull-type dogs were, however, overrepresented among dogs returned multiple times and had notably higher odds of euthanasia after a return.

Does a dog’s age matter more than its breed when it comes to being returned?

Yes, by a wide margin in the available data. Adult dogs between 2 and 8 years old had 3.4 times the return odds of puppies — a far larger effect than any breed-group comparison in the same study.

What’s the single most common reason dogs get returned to shelters?

Behavioral issues, cited in 36.1% of dog returns in the Charleston Animal Society data, followed by incompatibility with an existing pet in the home at 18.3%.

Does removing breed labels in shelters actually change outcomes?

In one controlled study, yes: dogs labeled “pit bull” took about three times longer to be adopted than similar-looking dogs without the label, and removing breed labels was associated with a 12% drop in euthanasia for that group.

Is this data representative of every shelter in the country?

Not necessarily — it comes from one shelter (Charleston Animal Society, South Carolina) over a five-year period. It’s the most rigorous data available on this specific question, but regional adoption patterns, shelter policies, and population differences mean the exact numbers could vary elsewhere even if the general patterns hold.

Bottom Line

The data doesn’t support ranking most breed groups against each other the way shelter-return listicles usually do — toy breeds and terriers have a real, measurable edge, but the rest of the field isn’t meaningfully different. What actually moves the needle is a dog’s age relative to your lifestyle, and for pit bull-type dogs specifically, the problem isn’t getting adopted once — it’s that the system treats a second chance far less generously, for reasons that are only partly about behavior.