Milan Krasula/Getty Dog attacks are on the rise in Australia. The most recent data from the Australian Institute of Health and Welfare show dog-related hospital admissions more than doubled in the eight years to 2021. Over 2021–22 in Australia, there were more than 9,500 cases of attacks where a person was bitten or struck by a dog. In response, some governments are introducing tougher penalties for dog attacks. In November 2025, for example, the South Australian government introduced fines of up to A$25,000 for people whose pet attacks and seriously injures or kills a person or animal. But reactive enforcement like this can’t prevent tragedies. Something that might be able to help is an Australian canine brain bank. This would be a key step in developing genetic tests that could identify high-risk animals before they ever bite.Breeds are complexIn March 2020, 90-year-old Ada Holland was killed by three unregistered dogs on Collingwood Beach in Vincentia, New South Wales. Her death prompted a coronial inquiry which heard that just a week before the fatal attack, the same dogs had attacked another person. However, inadequate council procedures failed to prevent the subsequent tragedy.During the inquest, one of us (Paul) provided expert evidence. Asked to identify the breed of dogs involved in the attack, he could only conclude they appeared to be a mixture of multiple dog breeds which included Staffordshire bull terrier; the specific breed composition could not be determined. Deputy State Coroner Carmel Forbes acknowledged breed-related issues remain “complex”. This uncertainty is a fundamental limitation of our current approach when it comes to identifying aggression in dogs, which relies on appearance, not neurobiology. The genetics of dog behaviourIn the past few decades scientists have made slow but steady progress in understanding the genetics of dog behaviour.A 2016 study found genetic variants linked to behaviours such as stranger-orientated and dog-orientated fear and aggression. Another study three years later identified genomic regions associated with aggression and fear across more than 100 breeds. Most significantly, a 2022 study demonstrated that breed alone explains less than 10% of behavioural variation.Our team’s pilot work has also shown increased serotonin 1A receptors in dogs euthanised for aggression. Genetic studies have linked serotonin and dopamine-related genes with canine aggression.These represent genuine advances in understanding aggression in dogs. A canine brain bank would go even further. What is a canine brain bank?Our research team first proposed establishing an Australian canine brain bank to understand the biological basis of impulsive aggression in 2013. But more than a decade later, it still doesn’t exist. The concept involves systematically collecting brain tissue from the thousands of dogs euthanised for aggression in Australia each year alongside normal controls and identifying receptor densities that characterise high-risk dogs. These patterns will then be linked to genetic markers predicting which dogs pose the highest risk – before they ever bite.Unlike probing a living dog’s brain, testing for the genetic markers predictive of impulsive aggression may eventually require only a simple blood sample. The brain bank essentially provides the key to translate neurobiological risk into testable genetic markers for living dogs. Dogs can lead double lives – loyal in familiar contexts, capable of sudden aggression when circumstances change. Connor Scott McManus/Pexels Building better scienceShelters already use behavioural assessments to evaluate how dogs respond to handling, food, and unfamiliar people. These assessments are useful, but they have limitations. For example, dogs can lead “double lives” – loyal in familiar contexts, capable of sudden aggression when circumstances change. A dog may pass shelter assessments, yet later respond dangerously to triggers not encountered during evaluation.A shelter could combine observed behaviour with genetic risk assessment. A dog showing good temperament but carrying high-risk genetic markers might need extra socialisation or placement only with experienced handlers. Conversely, a dog breed perceived to be dangerous but with low genetic risk markers might be safely rehomed.This combined approach could target resources where most needed. It could also help breeders select away from high-risk traits and give shelter staff objective assessment tools. But it’s important to say that genetic markers should only inform management strategies, such as placement with experienced handlers or enhanced training, and not mandate euthanasia decisions.Nature versus nurtureOf course, genetics and behavioural assessment don’t tell the whole story. Each dog’s learning history and gene expression – including their socialisation experiences, training methods, living conditions and how the owner handles the dog – profoundly shape behavioural outcomes. Even dogs with genetic predispositions toward impulsivity may never display dangerous aggression if properly managed, while dogs with low genetic risk can become dangerous through neglect or deliberate conditioning. This doesn’t diminish the value of genetic markers but underscores why we need multiple layers of assessment. Genetic information would help identify which dogs need particularly careful environmental management and which owners need additional support to provide it.No single approach will prevent all attacks. But better science – including the establishment of a canine brain bank – gives us much better odds.Paul McGreevy has received funding from the Australian Research Council, RSPCA Australia and animal welfare focussed philanthropy. He is a fellow of the Royal College of Veterinary Surgeons, a member of the British Veterinary Association and currently sits on the NSW Veterinary Practitioners Board.Rimini Quinn receives funding from a Sydney University scholarship and is a member of Animals Australia and VALE. She has worked professionally and in voluntary roles for companion animal rescue organisations RSPCA and Sunshine Coast Animal Refuge Society (SCARS).