Dog Bite Statistics by Breed: CDC, AVMA and Insurance Data

If a breed was named in a news story, a lease, an insurance form or a city ordinance, you want the real dog bite statistics by breed and a straight answer about what they prove. The United States has counts of bites, injuries, deaths and insurance claims, and some of those counts name breeds, but no reliable per-breed bite rate exists. This page gives every figure with its publisher and year, explains what each source can and cannot say, and shows how to use the numbers without misreading them.

Dog Bite Statistics by Breed: What the Data Actually Shows

People who look up breed numbers usually expect a per-breed bite rate, and that rate does not exist. The accurate answer is this: the data can show which breeds appear most often in certain records, but it cannot show which breeds are most likely to bite. A bite rate needs bites by a breed and dogs of that breed. The first number is partial and usually guessed by eye. The second does not exist, because no national registry counts dogs by breed, no dog breed population statistics exist beyond registered purebreds, and most dogs are never licensed. Unknown breed populations mean there is no reliable per-breed bite rate. The American Veterinary Medical Association (AVMA) says so directly: it is not possible to calculate a bite rate for a breed or to compare rates between breeds, because the breed of a biting dog is often unknown or misreported, most bites are never counted, and nobody knows how many dogs of each breed live in a community.

From ChewyIn stock

Embark Breed & Health Dog DNA Test Kit

$139

4.5

Buy on Chewy

Petful may earn a commission when you click through to Chewy, at no extra cost to you.

The numbers on this page are therefore organized by source rather than by breed. The source comparison below shows what each publisher counts, its date window, how breed is identified, who compiled it, and what each source cannot tell you. Every figure carries a publication year and a publisher link so you can check it.

Reading the table: only two rows name breeds at all, and both rely on labels from reporters, witnesses or photographs rather than pedigree or DNA. When a page says a breed causes a percentage of "dog bites," the figure almost always came from a fatality table covering a few dozen deaths a year. There is no official dog bite statistics by breed list, no measured dog breed with most bites per year, and no national count of unprovoked dog attacks by breed.

Which dog breeds have the highest bite rate?

No dog breed has a measured bite rate, because there is no national count of dogs by breed to divide bites by. The breeds that appear most often in the records are pit bull-type dogs, Rottweilers, mixed-breed dogs and German Shepherd Dogs, in that order in the fatality tables above (the CDC and AVMA study for 1979 to 1998 and the DogsBite.org tally for 2005 to 2019), and German Shepherd Dogs and mixed-breed dogs in the far larger set of hospital and animal-control bite records; those are counts, not rates. Most of those breed labels were assigned by eye, and DNA studies show visual identification is often wrong. The AVMA's review notes that controlled studies comparing biting dogs with the local dog population do not single out pit bull-type dogs. Different records, different questions, and none is a rate.

From ChewyIn stock

Wisdom Panel Essential Breed Identification DNA Test for Dogs

$95.99

4.5

Buy on Chewy

Petful may earn a commission when you click through to Chewy, at no extra cost to you.

How Many Dog Bites Happen in the US Each Year?

About 4.5 million dog bites happen in the United States each year, and more than 800,000 of those bitten need medical attention, according to the AVMA, citing the CDC; at least half are children. The AVMA repeated the 4.5 million figure in its April 2026 National Dog Bite Prevention Week release. Neither number carries a breed breakdown. There are no CDC dog bite statistics by breed for any recent year and no national dog bite statistics by breed from any federal agency; the CDC dog bite statistics that exist measure people, injuries and deaths. National dog attack statistics for the USA come from these CDC and AVMA estimates.

The CDC's current guidance, reviewed in June 2026, frames severity:

• Nearly 1 in 5 people bitten requires medical attention.
• Children are more likely than adults to be bitten, and their injuries tend to be more severe.
• Most bites to young children happen during everyday activities with familiar dogs.
• Adults living with two or more dogs are five times more likely to be bitten than people without a dog.

Three federal datasets add the hospital picture, and each carries its year:

• Emergency visits, 2001: the CDC's MMWR report estimated 368,245 people treated in emergency departments for nonfatal dog bites, 12.9 per 10,000 people, with the peak rate in children ages 5 to 9.
• Emergency visits and hospital stays, 2008: the AHRQ Healthcare Cost and Utilization Project brief counted about 316,200 emergency visits (103.9 per 100,000) and 9,500 hospital stays (3.1 per 100,000), about 866 visits and 26 admissions a day, with an average stay costing 18,200 dollars.
• Deaths, 2011 to 2021: the CDC's 2023 MMWR QuickStats counted 468 deaths from being bitten or struck by a dog, an average of 43 a year, ranging from 31 in 2016 to 81 in 2021.

The breed conversation online is built almost entirely on that last, smallest slice. Bites that look minor can still turn serious, and Petful's guide to why dog bites are more dangerous than they look covers infection risk and when to see a doctor.

Fatal Dog Attacks by Breed: What the Data Shows and Its Limits

Fatal dog attack statistics, the dog bite death statistics by breed that circulate online, are the only US figures that name breeds systematically, and they are the figures most often misused. Two tabulations exist. The first is the study by Sacks and colleagues, published in JAVMA in 2000 by CDC and AVMA authors, which combined news accounts with a Humane Society of the United States registry to identify fatalities from 1979 through 1998. The second is the ongoing count by DogsBite.org, a victims' advocacy group that supports breed-specific laws, which finds cases through news alerts and public-records requests. The table below tabulates deaths by the breed each source reported, side by side from the CDC and AVMA study for 1979 to 1998 and the DogsBite.org count for 2005 to 2019, with how breed was identified.

The JAVMA study reports that at least 25 breeds were involved in 238 deaths over 20 years and that pit bull-type dogs and Rottweilers were involved in more than half. In the same abstract the authors add that other breeds may bite and cause fatalities at higher rates, that a dog's breed is hard to determine with certainty, and that fatal attacks are too small a share of bites to drive public policy. The AVMA's later statement went further: the study cannot be used to infer breed-specific risk without knowing how many dogs of each breed exist.

The DogsBite.org tally is a media-report tabulation supplemented by records requests. Its data collection method page explains that cases are found through Google News alerts, that up to 33 parameters are collected per case, and that breed identification photographs were obtained in 81 percent of cases during one tracked year. Its 15-year fatality chart, the source of every DogsBite.org breed row in the table above, totals 522 deaths for 2005 to 2019, close to the CDC's death-certificate total of 519 for the same years, which supports the count of deaths. The breed shares rest on visual identification by reporters, officers and the group itself, and its own chart treats pit bull as a type covering several breeds and mixes. For 2020 onward the group reports its counts as still under review, so this page cites no per-breed shares for those years.

Where the CDC Stands on Breed Data

The CDC's position is the third reading. After the 1979 to 1998 study the CDC stopped collecting breed data on fatal attacks, and a CDC epidemiologist told the AVMA's 2001 convention that no centralized bite reporting system exists and that identifying breed reliably was one of the enormous difficulties in the data, as reported in AVMA's JAVMA News. Put together with the AVMA's finding that breed is not a reliable predictor, the fatality tables show which labels were written on death reports, not which breed is dangerous.

The pit bull label lumps American Pit Bull Terriers, American Staffordshire Terriers, Staffordshire Bull Terriers and their mixes together, and no source records non-fatal pit bull injuries nationally; death by dog attack statistics are the only breed-labeled numbers that exist.

From ChewyIn stock

Pupford Train Chicken Natural Freeze-Dried Dog Treats, 4-oz bag

Freeze-dried chicken training rewards with a compact format for rewarding individual repetitions. Check the label for ingredients and feeding directions.

$16.85

Buy on Chewy

Petful may earn a commission when you click through to Chewy, at no extra cost to you.

What breed of dog has the worst bite?

By injury severity, large powerful breeds cause the most serious injuries: pit bull-type dogs and Rottweilers are the breeds most often identified in fatal attacks in the JAVMA study (1979 to 1998) and the DogsBite.org tally (2005 to 2019), while small breeds bite more often, according to the AVMA's review of owner surveys, but cause less damage and rarely reach hospital data. The JAVMA authors caution that fatality counts cannot show whether those breeds bite more often or more severely per dog. The AVMA's literature review adds that in the small number of very severe or fatal cases pit bull-type dogs are more frequently identified, but that this may reflect the breed's popularity in the victim's community, reporting bias and how the owner treated the dog; in parts of Canada, fatal attacks are attributed mainly to sled dogs and Siberian Huskies because those are the dogs present. Size matters more than label: larger dogs of any breed are implicated in more attacks. Comparisons of pit bull statistics vs other dogs fail because a percentage of deaths compared with nothing is not a comparison, and there are no pit bull aggression statistics in these records, because aggression is measured by behavioral assessment, not injury reports. Dog mauling statistics by breed describe outcome severity in a handful of cases, not a bite-strength ranking, which this page does not attempt.

Non-Fatal Bites: Which Breeds Appear in Records and Why

Non fatal dog bite statistics by breed come from hospital charts, animal-control bite reports and rabies post-exposure logs, usually from one city or hospital over a few years. The AVMA literature review summarizes dozens of these studies from 1971 onward and finds the breeds most often listed are the German Shepherd Dog and mixed-breed dogs, followed by pit bull-type dogs, Rottweilers, Jack Russell Terriers and, in individual studies, Chow Chows, spaniels, collies, Saint Bernards and Labrador Retrievers. The review's explanation is popularity: when American Kennel Club registrations of Rottweilers peaked between 1990 and 1995, Rottweilers rose to the top of hospital bite lists in the late 1990s and early 2000s for the first time. Single-breed searches such as german shepherd attack statistics, golden retriever bite statistics or rottweiler bite statistics need that context:

• German Shepherd Dogs appear at or near the top of more bite studies than any other breed, including a 1991 Denver-area controlled study that compared biting dogs with matched non-biting households and found German Shepherd Dogs and Chow Chows disproportionately involved. The review calls the breed's results mixed, suggesting high variability by region and ownership.
• Golden Retrievers and Labrador Retrievers are notably under-represented relative to how common they are, although the review notes known aggressive subtypes. In a 2013 series of 537 children treated for facial dog bites, Labradors were the second most common breed named after mixed breeds, at 13.7 percent, which reflects how many Labradors live with children rather than a breed tendency.
• Rottweilers, Dobermans, Great Danes, Akitas, Huskies and mastiffs show up in individual hospital studies as large dogs whose bites are more likely to need treatment; larger dogs of any breed cause more of the bites that reach a hospital.
• Small dogs are the reverse. Behavioral assessments and owner surveys in the review find that the small dog breeds that bite the most, along with medium breeds such as collies, toy breeds, spaniels and, in one Canadian survey, the Lhasa Apso, Springer Spaniel and Shih Tzu, are more likely to bite than large breeds, yet they rarely appear in injury records because their bites rarely need a doctor. That is why small-breed bites are largely missing from the injury data.
• Breeds people search one at a time, such as Poodles, Dalmatians, Beagles, Corgis, Border Collies, Jack Russells, Bull Terriers, Bulldogs, Staffordshire Bull Terriers, Dogo Argentinos and wolf-dog crosses, appear only when a particular study logged them; a Jack Russell Terrier ranks in the top two in three British hospital studies, a Dalmatian, Beagle or Corgi tops none.

In the pediatric facial-bite series, nearly 90 percent of the dogs were known to the child and just over half of the bites were provoked by the child's own actions: a familiar dog, a young child, an everyday interaction.

What breed of dog bites the least?

No breed is bite-free, and no record can name the breed that bites least, because low counts mostly mean small populations, small mouths or under-reporting rather than gentle temperament. The AVMA's review is explicit that the breeds most likely to bite in owner surveys are small breeds that almost never appear in hospital data, so under-reporting of small dog bites is built into every table. The breeds genuinely under-represented relative to their numbers are the large retrievers. The question behind "bites the least" and "most aggressive" is temperament, not incidence, and it belongs to Petful's guide to the most aggressive dog breeds, which works from behavioral assessments, breed histories and owner surveys and explains why aggression varies more within a breed than between breeds. Nothing in bite counts can produce a top-ten list of aggressive breeds.

Why Breed Statistics Mislead: Breed Identification and Missing Denominators

Two problems sit under every breed table on this page. The first is breed identification. Nearly all breed labels in bite and fatality records were assigned by looking at the dog, and DNA studies show how often that goes wrong. In a 2013 study led by Victoria Voith, 923 dog professionals watched videos of 20 dogs; for 14 of the 20, fewer than half named any breed that matched the dog's DNA, and for only 7 dogs could even half the observers agree on a predominant breed. In a 2015 study of 120 shelter dogs in Florida, DNA identified 21 percent as pit bull-type while shelter staff collectively labeled 52 percent that way, and one in three dogs with no pit bull heritage was labeled a pit bull by at least one staff member.

From ChewyIn stock

Frisco Traffic Leash with Dual Padded Handles & Poop Bag Dispenser, Black, 6-ft long, 1 count

$14.99

4.7

Buy on Chewy

Petful may earn a commission when you click through to Chewy, at no extra cost to you.

Apply that error rate to a fatality table and "66 percent pit bull" becomes "66 percent of dogs that someone described as a pit bull," a category the American Kennel Club does not recognize as a breed and one the AVMA's review says cannot be reliably identified, with witnesses predisposed to assume a vicious dog is of this type.

The second problem is the missing denominator. The CDC and AVMA fatality study could not compute breed-specific risk because reliable breed population data do not exist: only a fraction of pet dogs are licensed, and those that are skew toward certain breeds. Registry counts cover registered purebreds and exclude most of the mixed-breed and unregistered dogs that fill bite records. Without that count, "pit bulls caused X percent of deaths" cannot become "a pit bull is X times more likely to kill."

For an owner, the practical side is a dog carrying a label it may not deserve on a shelter card, a lease application or an insurance form. If a label is affecting your housing or coverage, a DNA breed test replaces a guess with a lab result you can show. The Embark Breed & Health Dog DNA Test Kit reports breed ancestry down to small percentages along with genetic health screening, the combination most people want if the result may be shared with a landlord, insurer or veterinarian.

What a test can and cannot do:

• It can replace a visual label with a documented ancestry breakdown, which some insurers and landlords accept when a policy or lease hinges on breed.
• It cannot predict whether your dog will bite; the AVMA review, quoting Duffy and colleagues, finds within-breed variation too large to predict an individual dog's aggression from breed alone.
• It cannot override a rule written around appearance, so read the exact wording of an ordinance or policy before relying on a result.

If ancestry is the only question, a breed-only kit such as the Wisdom Panel Essential Breed Identification DNA Test for Dogs covers it at a lower price, and its report is the same kind of evidence when a label is challenged.

Dog Bite Insurance Claims and Breed Restrictions

Insurance data is the one breed-related dataset updated every year, and it measures money rather than bites. Each April the Insurance Information Institute (Triple-I) and State Farm publish the previous year's homeowner and renter liability claims for dog-related injuries, which include bites plus knockdowns and other injuries caused by dogs. The Triple-I dog bite liability spotlight, updated with 2025 data in April 2026, covers homeowner and renter insurance claims: claim count, total paid and average cost per claim.

• 28,450 claims in 2025, up 25.6 percent from 22,658 in 2024 and up 57 percent over the decade from 18,123 in 2016.
• About 1.86 billion dollars paid in 2025, up from about 1.57 billion in 2024 and 602 million in 2016, a 209 percent rise in ten years.
• An average cost per claim of 65,450 dollars, down 5.5 percent from 69,272 dollars in 2024 but nearly double the 2016 average of 33,230 dollars; New York had the highest average at 92,154 dollars.
• California led with 2,830 claims and Florida followed with 2,347; Michigan, Ohio and Pennsylvania rounded out the top five.

State Farm's own newsroom reports that in 2022 it paid nearly 211 million dollars on more than 3,285 dog-related injury claims, per its April 2023 release. None of these totals is broken out by breed, and State Farm does not ask about dog breed or type when writing homeowner, renter or business policies.

Many other insurers do restrict breeds and may decline coverage, exclude dog liability or charge more. Their lists usually name:

• pit bull-type dogs
• Rottweilers
• Doberman Pinschers
• Chow Chows
• Akitas
• mastiffs
• wolf hybrids and a few others

Which companies restrict which breeds changes often, so the live Petful insurance company dangerous dog list tracks it by insurer rather than this page repeating it. Triple-I notes that standard homeowner and renter policies typically cover dog bite liability up to the policy limit, commonly 100,000 to 300,000 dollars, and because restriction lists are built on breed labels, a dog labeled by a shelter's guess can be declined for a breed it may not be. The same fatality tables feed breed-specific legislation, which the AVMA opposes because no study has shown a ban reduces the rate or severity of bites; Petful's piece on how breed bans affect families covers that debate.

Who Gets Bitten: Dog Bite Statistics by Age

Age is the variable the federal data measures well, and it points prevention somewhere more useful than breed. Children are bitten most, older adults are hurt worst, and both are usually bitten by a dog they know. The figures below come from the CDC and AHRQ; any breed-and-child figure can only come from the fatality tabulations, with every identification problem described above. Prevention starts with the puppy: WebVet's vets explain how to handle hard puppy biting and growling before it becomes an adult habit.

• Children ages 5 to 9 have the highest emergency visit rate. The 2008 HCUP brief put it at 199.3 visits per 100,000 for ages 5 to 9 and 175.0 for children under 5, against 103.9 for the whole population; the CDC's 2001 surveillance found the same peak age.
• The youngest children are bitten on the head and neck. In the CDC's 2001 data, 64.9 percent of injuries in children under 4 were to the head or neck, a share that fell with age; the AVMA's 2026 release repeats that severe bites to young children are more likely to involve the head and neck.
• Children are bitten by dogs they know. The CDC reported that 80 percent of bites to people under 18 came from a family dog (30 percent) or a neighbor's dog (50 percent).
• Older adults are bitten less often but hospitalized more. In 2008, people 65 and older were 7.3 percent of dog bite emergency visits but 18.2 percent of hospital stays, and the hospitalization rate for ages 65 to 84 (4.5 per 100,000) was the highest of any age group.
• Deaths have shifted toward adults. CDC death counts more than doubled for both men and women between 2018 and 2021, and the DogsBite.org 2005 to 2017 report, whatever its breed limits, records 52 percent of victims at age 10 or older, with 76 deaths among people 70 and up.

Because the victim data is this consistent, here is prevention advice tied to victim-age data:

• Toddlers and preschoolers: never leave them alone with any dog, including the family dog. There are no separate national statistics on dogs attacking babies; infants sit inside the young-children group above, so the rule is the same for a newborn as for a toddler.
• School-age children: teach them to ask before petting and to leave eating or sleeping dogs alone.
• Older adults: prevent jumping and pulling that cause falls, and treat punctures promptly.

For walks and yard time around toddlers, a leash with a short second handle near the collar, such as the Frisco Traffic Leash with Dual Padded Handles, lets you keep the dog snug at your side the moment a child approaches.

Reward-based socialization teaches the other half: pay the dog for sitting calmly while a child greets it, using a small, soft treat like Pupford Train Chicken Freeze-Dried Dog Treats so calm greetings become the habit.

Petful's guide to the warning signs a dog may bite covers the body language both groups need to recognize.

Used this way, dog bite statistics by breed stop being an argument about which dogs to fear and become what the evidence supports: a case for supervising children, managing individual dogs, checking what a breed label is actually based on, and treating bites early.

Dog Bite Statistics FAQ