Breed averages in questionnaire studies do differ, but they are group averages, and they say very little about the one dog on your couch or the one you are thinking of bringing home. I'm not going to hand you a list of safe and dangerous breeds, and nothing here is a breed rule. The questions that actually help are what this dog does, what sets it off, and what its health and history look like.
Key takeaways
Hover or tap a card to bring it forward.
- Breed averages differ in questionnaires, but breed explains about 9% of behavioral variation between dogs, so it cannot predict your dog.
- Breed labels on shelter dogs and breed counts in bite records are unreliable, so neither makes a good rule.
- Size, sex, age and fear all show up in reported aggression, as associations and not proven causes.
- Same-sex fights show up often in clinic caseloads, but age gap, resources, management and temperament matter, so judge the actual pair.
- Judge the dog by its history and behavior, see a vet first for new or worsening aggression, manage the triggers, and get professional help after a bite or repeated fights.
What the breed-comparison studies measured
Start with the best-known comparison, because it is where most of the "some breeds are worse" talk comes from. Duffy, Hsu and Serpell used the C-BARQ, a validated owner questionnaire, in 2008. They analyzed two samples: 1,521 dogs from 11 breeds recruited through breed clubs, and 3,791 dogs from 33 breeds in an online sample, after mixed or unknown dogs and breeds with fewer than 45 responses were removed. The questionnaire asked about aggression toward unfamiliar people, familiar people, unfamiliar dogs and dogs in the same household.
Breed differences were statistically significant for stranger-directed, owner-directed and dog-directed aggression. Aggression toward unfamiliar dogs was reported more often than aggression toward unfamiliar people, and both were reported more often than aggression toward household members. In the online sample, owners of more than 20% of the Akitas, Jack Russell Terriers and Pit Bull Terriers reported snapping, biting or attempting to bite an unfamiliar dog. Those are findings about that sample, not advice about any dog you meet.
The authors also reported wide variation inside every breed. They flagged the limits themselves: owner-reported breed labels, a self-selected online sample, unequal group sizes and questionnaire answers instead of independently observed behavior. The results show an association between breed group and reported behavior. They do not show that breed caused it, and they say nothing about population-wide bite rates.
Then there is Morrill et al., published in Science in 2022. They surveyed the owners of 18,385 purebred and mixed-breed dogs and genotyped 2,155 of them. Breed explained about 9% of the behavioral variation between individual dogs. Some behavioral traits had heritable components, and the team identified 11 genetic locations tied to behavior. But no trait showed up in every dog of one breed or was missing from every dog of another, and the authors concluded that breed is not a reliable predictor of an individual dog's behavior. That 9% covers behavior overall. It is not an estimate of how much genetics contributes to biting.
So, the honest answer to "is it genetic?" is that some of it is, a modest share of it tracks breed on average, and none of that tells you what your dog will do.
Hover or tap a row to highlight it.
| Study | Who was in it | What it found | Limit to keep in mind |
|---|---|---|---|
| Duffy, Hsu and Serpell, 2008 | 1,521 dogs from 11 breeds and 3,791 dogs from 33 breeds | Breed differences in aggression toward strangers, owners and other dogs; in the online sample, owners of more than 20% of Akitas, Jack Russell Terriers and Pit Bull Terriers reported snapping, biting or attempting to bite an unfamiliar dog | Owner-reported labels and answers, a self-selected online sample and wide variation inside every breed, so it shows association, not cause |
| Morrill et al., 2022 | 18,385 dogs surveyed, 2,155 genotyped | Breed explained about 9% of behavioral variation between dogs; 11 genetic locations were tied to behavior | No trait appeared in every dog of one breed or was missing from every dog of another |
Why breed labels and bite statistics mislead
Part of the trouble is that the breed label itself is shaky. Gunter, Barber and Wynne genetically tested 919 shelter dogs in Phoenix and San Diego. Only 4.9% were purebred. At the San Diego shelter, staff matched at least one breed in a dog's genetic makeup in 67.7% of cases, but matched more than one breed only 10.4% of the time. A dog is usually a blend, and an eyeball guess captures a slice of it.
Olson et al. compared shelter staff's visual identification of Pit Bull-type dogs with DNA profiles. Sensitivity ranged from 33% to 75% and specificity from 52% to 100%, depending on the person doing the identifying. Some dogs with relevant ancestry were missed, and some dogs without it were labelled as Pit Bull-type.
Bite and fatality records have a different problem. They list which dogs were reported in particular incidents, but they have no reliable denominator showing how many dogs of each breed were living in the area. They are also shaped by how much contact each type of dog has with people, size and strength, how incidents get reported, media attention, owner behavior, injury severity and that same uncertain breed identification. The American Veterinary Medical Association says whether a dog bites is influenced by its individual history and behavior more than by breed alone. A bite record describes an incident. It does not turn a breed into a rule.
Hover or tap a row to highlight it.
| Source | What was checked | Result |
|---|---|---|
| Gunter, Barber and Wynne | 919 shelter dogs in Phoenix and San Diego tested genetically | Only 4.9% were purebred |
| Gunter, Barber and Wynne | San Diego shelter staff guesses | At least one breed matched in 67.7% of cases, more than one breed matched only 10.4% of the time |
| Olson et al. | Staff identification of Pit Bull-type dogs versus DNA | Sensitivity from 33% to 75%, specificity from 52% to 100%, depending on the person |
| Bite and fatality records | Which dogs were reported in incidents | No reliable count of how many dogs of each breed live in the area |
Size, sex and neutering
Some of what gets blamed on breed may be about body size, sex or age. A Finnish owner questionnaire covered 13,715 purebred dogs from 264 breeds. Small dogs had a higher probability of owner-reported aggression toward people than medium and large dogs. Male dogs, older dogs and fearful dogs also had higher reported rates, and small size was linked to fear-related behavior too. It is an association, not proof that being small causes aggression. Fear when approached by bigger people or dogs, differences in how small dogs are handled, and how owners react could all play a part. A small dog that lunges often gets picked up; a large one that lunges gets noticed.
Roberts et al. analyzed C-BARQ data from 8,301 dogs across 49 breeds and linked height, weight and skull shape to several owner-reported behavior traits, including aggression-related ones. The sample was breed-based and questionnaire-based, so it is not a universal size rule.
Neutering questions come up constantly, so here is what Farhoody et al. reported from C-BARQ data on about 13,796 dogs. After accounting for several variables, gonadectomy was not linked to aggression toward familiar people or toward other dogs. Dogs neutered at 7 to 12 months showed a small increase in aggression toward strangers. The data were observational and neutering was not randomly assigned, so nobody can say the surgery caused the difference. Sex, age, reproductive status, fear, learning history and experience each deserve their own look rather than being lumped together.
Hover or tap a row to highlight it.
| Factor | What the studies reported | Keep in mind |
|---|---|---|
| Small size | Higher probability of owner-reported aggression toward people than medium and large dogs (Finnish study, 13,715 dogs from 264 breeds) | An association, not proof that being small causes aggression |
| Male, older and fearful dogs | Higher reported rates in the same Finnish study | Fear and handling may play a part |
| Height, weight and skull shape | Linked to several owner-reported traits, including aggression-related ones (Roberts et al., 8,301 dogs, 49 breeds) | Breed-based questionnaire sample, not a universal size rule |
| Neutering | Not linked to aggression toward familiar people or other dogs; small rise toward strangers in dogs neutered at 7 to 12 months (Farhoody et al., about 13,796 dogs) | Observational data, so cause cannot be shown |
Two dogs of the same sex in one house
If you are adding a second dog, you have probably heard that two females, or two males, are a bad match. There is a real signal behind that, with limits that matter.
Sherman, Reisner, Taliaferro and Houpt reviewed 99 cases of dog-to-dog aggression treated at Cornell University from 1983 to 1993. Seventy-three involved dogs in the same household, and same-sex pairs were much more common than opposite-sex pairs, especially female pairs. A later veterinary behavior clinic study covered 305 pairs, or 610 dogs, seen between 2007 and 2016 for fighting at home. At least one female was in 214 pairs, or 70%, and 188 pairs, or 61.6%, were same-sex. Resource guarding was the most common trigger, in 222 pairs, or 72.8%. Same-sex pairing, skin-breaking bites and attacks the moment one dog saw the other were linked to poorer outcomes.
Here is the limit. These are referral caseloads, the homes where things went badly enough to need a specialist. They are not random households, so they cannot tell you that any particular pair is unsafe. Age gap, which dog arrived first, what the dogs compete over, how well you manage the house, and each dog's temperament all mattered in those cases. If you are choosing a second dog, ask about the actual dogs: history, triggers, age, arrival order, and whether they can live safely together with gates and separate feeding if needed.
Hover or tap a row to highlight it.
| Finding | Result |
|---|---|
| Cornell cases involving dogs in the same household | 73 of 99, with same-sex pairs much more common, especially female pairs |
| Pairs in the veterinary behavior clinic study | 305 pairs, or 610 dogs, seen from 2007 to 2016 |
| Pairs with at least one female | 214 pairs, or 70% |
| Same-sex pairs | 188 pairs, or 61.6% |
| Resource guarding as the trigger | 222 pairs, or 72.8% |
| Linked to poorer outcomes | Same-sex pairing, skin-breaking bites and attacks the moment one dog saw the other |
What shapes behavior besides breed
Early experience carries real weight. A review of early experience and human-directed aggression by McMillan reported that, in some studies, dogs bought from pet shops were up to 2.4 times more likely to show owner-directed aggression than dogs from breeders, and dogs from hobby breeders up to 1.3 times more likely than dogs from other breeder sources. Those figures come from observational work and may reflect breeding conditions, early handling, maternal care and later owner differences. A North American survey of puppy owners looked at socialization in puppies under 20 weeks old, and the authors concluded that some puppies get few early opportunities and that vet education on early socialization and reward-based training could help reduce later problems. Owner reports cannot prove which practice prevents aggression.
Training and handling matter too. Hiby, Rooney and Bradshaw surveyed owners about training methods. Puppy classes were linked to fewer undesirable reactions to dogs outside the household, and owners who only trained informally at home without classes were more likely to report aggression. Owners may have picked methods in response to problems they already had, so the direction of cause is open. Rooney and Cowan filmed 53 owners and their dogs: dogs trained with more aversive methods were less likely to join in play with strangers or with their owners. It is a small observational study that cannot separate method from owner personality or dog temperament.
The American Veterinary Society of Animal Behavior favors reward-based training and states that aversive methods carry welfare and relationship risks. My own approach is balanced and includes e-collar work among other tools. I tie any method to the dog, the handler and the protocol, and no method guarantees an outcome for an individual dog.
For the wider list of what drives a dog to aggression, I put together a page on the causes of dog aggression that goes past breed.
What to do instead of reading the breed
Judge the dog in front of you. The AVMA and the RSPCA both point to history and behavior, and the RSPCA says future aggression cannot be predicted with certainty for a dog with no known history. For a dog already in your home, ask four things. What happens just before the behavior? Who or what is the dog responding to? What does the dog gain by it? Has the pattern changed? Learning to read the early warnings helps, and a page on the signs of aggression in dogs walks through them.
If you are thinking of adopting or buying, watch the dog across more than one meeting. See how it behaves around unfamiliar adults, children, other dogs, food, toys, handling and confinement. Ask for any bite or fight history, medical records and the reason the dog was surrendered. A single shelter test or a brief greeting is not a forecast.
Vet first when aggression is new or getting worse. The Merck Veterinary Manual lists pain, neurological disease, organ dysfunction and hormonal disorders as possible contributors. Other medical causes include dental or skin disease, endocrine and metabolic problems, fever, medication effects, and cognitive change in older dogs. A veterinary review notes that the thyroid link to aggression is controversial, so thyroid disease should not be assumed from the behavior alone. Tell your vet when the change began, what triggers it, and any shifts in sleep, appetite, thirst, mobility or medications.
Management comes next. Use gates, closed doors, leashes, distance and trigger avoidance so the dog stops rehearsing the behavior, and keep a dog with a bite history out of situations where another bite is foreseeable. The RSPCA advises avoiding the triggers in the short term while professional help is arranged, and it warns that aggression can escalate with repetition. After that comes desensitization, showing the trigger at a distance or intensity the dog can handle, and counter-conditioning, pairing the trigger with something the dog values, adjusted by what the dog does. The RSPCA also warns that punishment during aggression can suppress warning signals or intensify the response, which is one more reason to read the warnings early.
Judging the Dog in Front of You
- Ask four questionsWhat happens just before the behavior? Who or what is the dog responding to? What does the dog gain by it? Has the pattern changed?
- See the vet firstWhen aggression is new or getting worse, a vet can check for pain, neurological disease, organ dysfunction and hormonal disorders. Tell the vet when the change began and any shifts in sleep, appetite, thirst, mobility or medications.
- Manage the triggersUse gates, closed doors, leashes, distance and trigger avoidance so the dog stops rehearsing the behavior, and keep a dog with a bite history out of situations where another bite is foreseeable.
- Add behavior workDesensitization shows the trigger at a distance or intensity the dog can handle. Counter-conditioning pairs the trigger with something the dog values.
- Bring in a professionalAfter a bite or repeated fights, get a behavior professional who takes a detailed history and gives you a structured plan with safety steps.
When to get help
Get help after a bite, repeated fights, injuries, aggression in several settings, a sudden behavior change, or severe fear. Aggression toward children belongs on that list; the page on dog aggression toward children covers the causes and what to do about them. Start with the vet, who can check for medical causes and refer you to a veterinary behaviourist.
A good behavior professional takes a detailed history, observes the dog safely, identifies triggers and warning signs, gives you a structured plan with safety steps, and coordinates with your vet. The RSPCA notes that anyone can call themselves a behavior expert, and AVSAB advises checking education, credentials and methods, asking the professional to explain exactly what they will do, and being wary of anyone who offers guarantees. Titles in this field are not protected, so ask questions.
Questions I get about breed and aggression
Are some breeds genetically more aggressive? Breed averages differ in owner questionnaires, and Morrill et al. tie about 9% of behavioral variation between dogs to breed. That leaves most of the difference between dogs coming from individual genetics, experience, health and environment.
Can a shelter tell me my dog's breed from how it looks? Not reliably. In Gunter, Barber and Wynne's San Diego shelter data, staff named more than one breed from a dog's actual genetic makeup only 10.4% of the time, so a label on a kennel card is a guess.
Does neutering make a dog more or less aggressive? In Farhoody et al.'s data, gonadectomy was not linked to aggression toward familiar people or other dogs, and the small rise in stranger-directed aggression appeared only in dogs neutered at 7 to 12 months. It was an observational study, so it cannot prove cause.
What should I do before adding a second dog? Ask about the specific dogs' history, triggers, ages and how they handle food and toys. Set up gates, separate feeding and supervised introductions first, because resource guarding was the trigger in 72.8% of the clinic pairs.
Check What You Picked Up
About how much of the behavioral variation between dogs did Morrill et al. link to breed?
In the San Diego shelter data, how often did staff name more than one breed from a dog's actual genetic makeup?
What was the most common trigger in the pairs of dogs fighting at home?
What should you do first when a dog's aggression is new or getting worse?
Why can a bite record not turn a breed into a rule?
Answer a question to see your score.
If your dog needs more than a page
If a bite or near-bite has already happened, say so on the first call, even if it felt small. I offer an evaluation visit, in-home lessons, and board and train in my own home for dogs whose problem needs more than advice from a page. When you are ready to talk it through, reach out through the contact page.