A dog races through the house, struggles to settle, barks at small noises, or appears unable to concentrate when another dog walks past. Someone may describe that dog as “neurodivergent.” The word can sound compassionate. It may communicate that the dog is not deliberately being difficult and that its experience of the world might differ from what people expect.
But does the label help us understand what is happening? More precisely, does it identify a scientifically established condition, explain the behavior's causes, or tell us what to do next?
These are not the same question. Dogs unquestionably vary in activity, attention, sensitivity, cognitive performance, brain anatomy, and responses to the world around them. Some dogs have genuine neurological disorders. Researchers are also investigating patterns they call ADHD-like or autism-like. Yet none of those facts, by themselves, establishes “neurodivergence” as a validated canine diagnosis.
The most useful question for a trainer, veterinarian, or dog owner is not merely what we can call a dog. It is what we can reliably observe, what might explain those observations, how we can distinguish competing explanations, and what information would improve the dog's welfare.
A popular word is not a scientific explanation
In human discussion, neurodivergent is a broad, non-diagnostic term for neurological or cognitive differences. It does not itself identify the specific biological cause of any one person's behavior. Clinical conditions such as ADHD and autism involve distinct human diagnostic frameworks. It would be a mistake to assume those frameworks transfer unchanged to another species.
Consider the difference between saying “my dog moves constantly” and “my dog has a neurological disorder that makes him move constantly.” The first is a report of behavior. The second proposes an explanation and a diagnosis. The first may be observable; the second requires evidence that is not contained in the description.
There are at least five levels we need to keep separate: what somebody reports; what can be independently measured; what stable characteristic, if any, we infer from those measurements; what biological or learning mechanisms might produce the behavior; and whether a legitimate clinical classification applies. Evidence at one level cannot automatically be promoted to the next.
A dog may behave differently from other dogs without those differences necessarily indicating a medical, neurological, or behavioral disorder. Conversely, a dog may experience substantial difficulties affecting its functioning or welfare without fitting any popular diagnostic label.
What the large canine behavior surveys actually show
One often-cited study by Sulkama and colleagues examined owner reports of hyperactivity, impulsivity, and inattention in more than 11,000 dogs. It found associations involving reported behavior and factors such as age, sex, breed, time spent alone, and other reported behavioral characteristics. These patterns are valuable: they can identify questions worth testing and help scientists study variation across large groups. [1]
They do not establish that those dogs had the human condition ADHD, nor that any behavior was caused by a particular brain abnormality.
This distinction is sometimes misunderstood as an attack on owners' observations. It is not. Owners frequently know when a behavior began, where it happens, who was present, and what has changed. Their accounts are indispensable. But two people may use the word hyperactive for different behavior, or report similar behavior under very different circumstances. A very large survey can estimate patterns in answers without eliminating ambiguity about what each answer means.
Suppose one owner describes a young dog that bounces around the kitchen while dinner is prepared. Another describes an older dog that paces for hours, sleeps poorly, and cannot settle even in a quiet room. Both may check the same box labeled “restless,” despite describing substantially different problems that call for different investigations.
A questionnaire can organize observations. It cannot replace the work of discovering what those observations represent.
Researchers are developing direct behavioral tests, but the distinction remains
This field has moved beyond surveys. In 2024, Csibra and colleagues studied 59 family dogs in structured tasks intended to measure attention, activity, and responses described as impulsive. The tasks included an attention exercise, a plush-dog encounter, a leash-related test, and a sit test. Some measures correlated with owner or trainer ratings; repeat testing of a smaller subset also showed meaningful consistency within certain tasks. Those are useful methodological developments. [2, 3]
The limitations matter just as much as the results. Participants were recruited partly in relation to questionnaire scores, the tested sample was small, and performance on a task is affected by a dog's experience, motivation, environment, and response to the particular test. Agreement between a test and a rating does not independently prove an underlying neurological disease.
In 2025, Csibra, Bunford, and Gácsi proposed a more formal combination of owner-rated symptoms and functional impairment. Among 1,872 dogs, 79 (4.22% of that study sample) met the proposed “at-risk” screening criteria. This is not evidence that 4.22% of the canine population has ADHD. The owners were not a representative census of all dogs, and the classification was a screening designation rather than a confirmed veterinary diagnosis. [4]
The authors reported an area-under-the-curve statistic, or AUC, of 0.861 for distinguishing questionnaire-defined impairment groups using symptom scores. An AUC can be informative within a research design. Here, however, the criterion and predictor came from related parts of the same owner-completed questionnaire. The number therefore does not establish independently verified clinical diagnostic accuracy against a separate veterinary gold standard.
The study's published title uses the phrase “human analogue ADHD diagnostic system.” That is the researchers' title, not a conclusion this article adopts. Their own paper identifies behavioral testing and expert consultation as further steps. For the present evidence, proposed behavioral screening approach is the more defensible description.
The impulsivity problem: are we even measuring one thing?
A particularly important challenge appears in a systematic review and meta-analysis by Barela, Worth, and Stevens. They identified 13 studies that had used more than one proposed impulsivity task with the same dogs. Across 31 comparisons between different tasks, 28 did not detect a statistically significant correlation. Among 15 comparisons involving owner ratings and task scores, 10 also failed to find a significant association. A meta-analysis of a repeatedly studied task pair found no overall correlation. [5]
This does not establish that dogs never display impulsive behavior. They plainly do things such as act before waiting for a cue or rush toward an immediate opportunity. The narrower scientific issue is whether different tests are tapping into a coherent, stable characteristic that predicts behavior across settings.
At present, the reviewed measures offer weak and inconsistent support for that broad interpretation. Some may primarily capture how a dog responds to a particular challenge, its reinforcement history, its interest in the test object, or its understanding of the task. Statistical nonsignificance is not proof that no stable individual difference exists anywhere, especially where tests and sample sizes vary. But it is also not a reason to assume the sought-after trait is real and merely waiting to be found.
The practical lesson is straightforward: a dog that rushes forward in one experimental task cannot automatically be characterized as globally impulsive, much less diagnosed with ADHD. Researchers must show which behaviors generalize, under what circumstances, and with what reliability.
Different brains? Yes. Evidence of a canine autism or ADHD brain signature? No.
There is strong evidence that canine brains are not anatomically identical. Hecht and colleagues used MRI scans from 62 dogs representing 33 breeds and identified regional anatomical differences associated with historical breed specializations. That research is important because it shows that the structure of the canine brain varies in nonrandom ways. But the behavioral specializations were largely characterized by breed history; the study did not establish a neurological diagnosis or independently test every dog's actual performance in the scanner. [6]
A more clinically suggestive study, by Ogata and colleagues, compared MRI scans from eight Doberman Pinschers with canine compulsive disorder and eight unaffected Dobermans. It reported group differences in several gray-matter measures and white-matter organization. Those are genuine observations made with an objective imaging method. They are not proof that particular brain structures caused the compulsive behavior, and the study was not a test of canine ADHD or autism. [7]
Why is causation difficult here? First, 16 dogs are a very small sample. Second, groups may differ in relevant characteristics besides the behavior being investigated. Third, the scans were obtained at one point in time. Without knowing what the brains looked like before the behavior appeared, we cannot confidently determine the direction of influence.
The nervous system changes with development, activity, stress, illness, and learning. A brain difference observed alongside years of repetitive behavior could be a contributing cause, a consequence of experience-dependent neural plasticity, or the result of another factor influencing both the brain and behavior. Genetic influences also deserve investigation, but an association with a gene does not by itself prove the pathway from a DNA variant to an individual dog's observed actions.
None of this diminishes the importance of neuroscience. It identifies what a brain scan can, and cannot, tell us without longitudinal and mechanistic evidence. An observed difference does not become pathological just because it is visible on an image.
What about cells, genes, and dogs said to have autism-like behavior?
Some of the most interesting research concerns persistent repetitive behavior, including tail chasing and circling. These behaviors deserve study because they may occur with genuine functional impairment and may have multiple interacting causes.
Tsilioni and colleagues compared blood measurements in Bull Terriers with persistent tail chasing and unaffected Bull Terriers. They reported differences involving serum neurotensin and corticotropin-releasing hormone, alongside comparisons with human research on autism-related biological findings. The dog sample in the published study included 14 tail-chasing and 18 unaffected Bull Terriers. Crucially, these were serum measurements, not microscopic observations of the dogs' brains. Overlapping biological findings across species can motivate hypotheses; they do not establish that the dogs had autism. [8]
Cao and colleagues examined dogs showing persistent circling, identified candidate genetic regions, and investigated how two candidate genes—PPP2R2B and ADAMTSL3—affected dendritic-spine characteristics in an experimental system. Dendritic spines are microscopic structures associated with neuronal connections. Such work contributes to understanding possible biological pathways, but it did not demonstrate that microscopic differences in the brains of individual circling dogs caused their behavior, or establish a canine autism or ADHD diagnosis. [9]
Here we should be especially careful about comparisons between species. A dog may exhibit a behavior that resembles one feature sometimes studied in human autism. A serum marker may move in a similar direction. A gene may be involved in neuronal development in both species. Those observations do not mean that the entire human clinical syndrome has been identified in dogs. Similarity is a research lead, not diagnostic equivalence.
Dogs with repetitive behavior deserve serious assessment rather than dismissal. The assessment should remain canine-specific and should consider medical, behavioral, developmental, environmental, and learning explanations rather than beginning with a borrowed human diagnosis.
A useful contrast: inherited canine narcolepsy
There are canine neurological conditions for which researchers have assembled much stronger causal evidence. In certain families of dogs with narcolepsy, scientists identified mutations affecting the hypocretin/orexin receptor 2 gene, HCRTR2. Additional work examined the effects of receptor mutations on function. In 2025, Mondino and colleagues described another familial variant involving a tandem duplication in HCRTR2 in Dogo Argentinos. [10–12]
This is an instructive example because multiple types of evidence converge: a reasonably well-defined disorder of sleep-wake regulation, inheritance patterns, gene-level findings, and functional investigation. It demonstrates what the careful connection of a biological mechanism to a specific canine condition can look like.
But the comparison has strict limits. Inherited narcolepsy in these families is a comparatively discrete neurological disorder linked to specific mutations. Complex patterns described as inattention, hyperactivity, repetitive behavior, or difficulty adapting are more heterogeneous. They may involve many genes, medical conditions, learning histories, and environments. Requiring one gene or one characteristic brain image as the only acceptable proof of every behavioral diagnosis would be just as mistaken as accepting every questionnaire score as one.
Nor should all forms of canine narcolepsy be assumed to have precisely the same genetic mechanism. The example illustrates successful causal investigation in specified familial cases, not a universal explanation for unusual canine behavior.
Nature versus nurture: why no single percentage will do
The question “Is it genetic or learned?” is understandable but often asks for an impossible partition. Genes influence the development and operation of a nervous system. Development occurs in an environment. Experience changes what animals learn, which situations they encounter, and how some neural circuits function. The resulting behavior cannot usually be assigned a single universal percentage to heredity and another to experience.
Morrill and colleagues reported that modern breed explained approximately 9% of individual behavioral variation in their large dataset. This does not mean genes explain only 9% and environment explains the other 91%. Breed designation is not identical to genetic heritability, and population-level variance is not the same as an explanation of an individual dog. [13]
In another important study, Lord and colleagues analyzed genetic and behavior data from more than 3,000 dogs. Candidate genetic variants previously promoted as behavioral predictors failed to show the expected predictive associations in that dataset. The study did not demonstrate that genetics is irrelevant. Rather, it showed the limits of simplified predictions from a few markers for complex behavior. [14]
Early experience also matters. Developmental work, including the classic experiments by Freedman, King, and Elliot, showed that the timing of exposure to people affected later social responses in their study conditions. A modern review by McEvoy and colleagues found useful evidence about canine socialization, but also substantial dependence on retrospective owner reports and a shortage of modern controlled experiments. [15, 16]
The better question is not whether genes or environment “win.” It is which inherited predispositions, physical conditions, experiences, and present-day contingencies matter to this dog, displaying this behavior, in this setting.
What learning theory can explain and what it cannot
Consider a dog that lies down on cue in a quiet kitchen but cannot remain lying down in a busy park. It is tempting to call the dog inattentive. Yet the park may contain moving animals, new smells, social uncertainty, noises, and opportunities that compete with the consequences of the owner's cue. The dog may not have learned the response sufficiently under those conditions. Different stimulus control is not, by itself, evidence of a neurological deficit.
Or consider a dog that barks at a faint sound outside. The sound could have become associated with visitors approaching, a startling event, or a previous pattern in which barking was followed by a person going away. Barking may be affected by arousal, sensory sensitivity, past learning, and current consequences. The observable behavior is real; its explanation still requires investigation.
Learning research provides tools for that investigation. Pavlovian learning helps explain how previously neutral events acquire significance. Instrumental learning concerns how actions change in relation to their consequences. Stimulus discrimination and generalization help explain why performance differs between settings. Research on extinction and renewal shows why a response can diminish in one context and reappear in another. [17]
We should not turn this into the opposite error and insist that every unusual response is “just training.” Pain, neurological illness, sleep disturbance, sensory impairment, medication effects, inherited characteristics, and other biological influences can alter an animal's opportunities to learn and perform. A learning explanation must make predictions and be tested, not simply replace one unsupported label with another.
A useful assessment therefore follows the interaction between the animal, the people around it, other animals, the physical setting, and the history of what happened before and after important events. The owner, trainer, and veterinarian may each see different parts of the same picture. Reconciling those observations can be more informative than choosing one label prematurely.
What about canine hyperkinesis? Isn't that a veterinary condition?
Veterinary references do discuss hyperkinesis, sometimes drawing comparisons with human ADHD. The Merck Veterinary Manual describes excessive activity, impulsivity, and related difficulties, while explicitly noting that the veterinary profession lacks consensus on diagnostic criteria for canine hyperkinesis. It also emphasizes differential explanations, including ordinary behavior when a dog's needs are unmet, anxiety, compulsive disorders, and problems with impulse control. [18]
This means two things at once. First, we should not claim that veterinarians have established a single, universally accepted canine ADHD diagnostic system comparable to that used in human medicine. Second, we must not pretend that clinically important canine behavioral and neurological disorders do not exist just because a particular popular term lacks validation.
A veterinarian may identify medical and behavioral disorders using clinical history, examination, and appropriate testing. A blood test or MRI is not required to legitimize every clinical diagnosis. Conversely, a trainer's impression, an online questionnaire, or a difficult training session cannot authorize a neurological diagnosis. Veterinary behavioral assessment must consider medical contributors, including pain, sleep and sensory problems, and relevant neurologic illness. [19]
How I would investigate an individual dog
If an owner tells me a dog is neurodivergent, I would first ask what prompted that description. Perhaps the dog struggles with frustration, repeats a particular action, reacts to sound, or behaves very differently around unfamiliar people. I would want examples rather than adjectives.
I would then reconstruct the circumstances: When did the behavior first appear? What is the dog's age and developmental history? Does it occur at home, outdoors, around particular people, or only in specific situations? What happened immediately before and after? Are there patterns in sleep, exercise, health, pain, medications, and previous training? How do different family members respond? What reliably makes the behavior better or worse?
A behavior log or brief video may be more useful than a broad label. If we identify a particular response, we can measure its frequency, duration, intensity, and setting, then assess whether a safe change in management or training affects it. If there are medical signs, a sudden change, persistent unexplained distress, or concerns suggesting a neurological condition, the appropriate next step is veterinary evaluation, not a trainer offering a diagnosis.
A dog may need a quieter starting environment, shorter practice sessions, a different reinforcer, more predictable routines, improved opportunities for rest, or an adjusted management plan. These are examples of questions to investigate, not treatments prescribed for a hypothetical neurological syndrome. The goal is to determine what actually helps the individual animal while protecting welfare and avoiding unnecessary risk.
What evidence would make a canine neurodevelopmental label more convincing?
A useful scientific classification must do more than rename a set of troubling behaviors. Researchers would need to define the proposed canine condition, show that different observers and instruments can identify it reliably, establish which features remain stable across appropriate settings, and demonstrate how it differs from plausible medical, developmental, and learning explanations.
Longitudinal studies would help determine whether a proposed biological difference precedes a behavior or follows it. Appropriate controls would reduce alternative explanations. Independent research groups would need to reproduce important findings. Genetic, imaging, physiological, and behavioral results would be stronger when they converge on predictions that cannot be made equally well by competing theories.
Clinical usefulness is another test. Does the classification improve assessment, predict the dog's needs or prognosis, or lead to more effective and humane decisions beyond a careful individual evaluation? A label that does not change what we can predict or do may offer limited scientific value even if it is memorable.
These requirements do not mean a single gene or unique biomarker must exist. Complex clinical conditions can be legitimate without such a marker. The point is that a proposed category should become more reliably distinguishable, explanatory, and useful as the evidence improves.
My conclusion
Dogs do not all think, learn, react, or develop in the same way. Some of those differences arise in brain structure and function. Some arise through developmental experience, health, learning, and changing circumstances. Most complicated behaviors reflect interactions among these influences rather than a clean choice between “nature” and “nurture.”
Research on canine ADHD-like ratings, behavioral testing, repetitive behavior, genetics, and neuroanatomy is valuable and deserves continued scrutiny. But the evidence reviewed here does not establish “neurodivergence” as a validated canine veterinary diagnosis, nor does it justify telling an owner that an unusually active, sensitive, or repetitive dog has human ADHD or autism.
The most useful response to an unusual dog is not to dismiss the behavior, romanticize it, or assign a fashionable name. It is to observe carefully, investigate competing explanations, recognize where veterinary expertise is needed, and adapt our expectations and training to what the individual dog shows us.
Over more than 30 years of working with dogs, I have encountered some truly puzzling behaviors. Some did not fit neatly into conventional behavioral categories. In other cases, even the dogs' veterinarians were unable to identify a satisfactory explanation.
These are the cases that remind me how much we still don't know.
Experience helps. Decades of observing dogs, working through behavioral problems, and seeing how different animals respond to different circumstances can provide valuable insights. Sometimes, however, experience and an educated hunch are all I have to go on. That doesn't make the hunch correct. It simply gives me a place to begin investigating.
The scientific study of animal behavior has made enormous progress, but many questions remain unanswered, and others have received surprisingly little research attention. That is part of what makes this field so fascinating. It can also make it extraordinarily frustrating.
I have seen owners turn to over-the-counter calming supplements, sometimes without noticeable improvement. In other cases, products or medications appear to suppress a dog's activity without necessarily addressing whatever is producing the behavior. I have also encountered dogs whose veterinarians prescribed medications but whose responses fell short of expectations. That does not mean those treatments are inherently ineffective or inappropriate. It means that identifying the right explanation and the right intervention for an individual dog is not always straightforward.
And we need to distinguish a dog that has become quieter or less active from a dog whose underlying distress or behavioral difficulty has improved. Those are not necessarily the same outcome.
Sometimes these cases require more investigation, additional veterinary evaluation, changes in the environment, or a different approach to training and management. Sometimes the available evidence simply does not allow us to reach a confident conclusion.
And sometimes we may never know exactly why a particular dog behaves as it does.
That uncertainty is uncomfortable, especially when people are looking to us for answers. But I believe it is better to acknowledge what we don't know than to provide an impressive-sounding explanation we cannot substantiate.
Ultimately, my concern is the welfare of the dog and the people and other animals affected by its behavior. Ideally, the dog is not suffering, its behavior does not create unreasonable hardship for its owners, and it does not present an unacceptable danger to anyone.
We may not always be able to explain a behavior with scientific certainty. But we can still work to understand the circumstances in which it occurs, reduce suffering, improve safety, and help the dog and its family live together more successfully.
Sometimes the responsible answer is not a diagnosis or a label. Sometimes it is simply an acknowledgment that we have a puzzle we haven't solved yet. The dog is real. The behavior is real. The explanation still has to be earned.
As I always say, “we can’t put a dog on the proverbial psychiatrist’s couch and ask them about their troubled childhood.”
Thus, we must continue to puzzle these things out. And hopefully research eventually answers these questions more definitively.
Research approach and limitations
This article is a focused, non-systematic narrative review using selected peer-reviewed canine studies, methodological literature, and professional veterinary guidance. Central claims were checked against accessible primary publisher, PubMed, or PubMed Central records as available through November 16, 2026, together with an internal evidence register. Some studies were assessed from full text or substantial accessible excerpts and others from abstracts or selected sections. This article is not a systematic review, raw-data reanalysis, or substitute for a veterinary assessment. Study claims are limited to what the accessible methods and findings can support.
References
[1] Sulkama et al. (2021). Canine hyperactivity, impulsivity, and inattention share similar demographic risk factors and behavioural comorbidities with human ADHD. Translational Psychiatry. https://doi.org/10.1038/s41398-021-01626-x
[2] Csibra, Bunford & Gácsi (2024). Development of a human-analogue, 3-symptom domain Dog ADHD and Functionality Rating Scale. Scientific Reports. https://doi.org/10.1038/s41598-024-51924-9
[3] Csibra et al. (2024). Towards an Objective Measurement Tool for ADHD-like Traits in Family Dogs: A Comprehensive Test Battery. Animals. https://doi.org/10.3390/ani14131841
[4] Csibra, Bunford & Gácsi (2025). Development of a human analogue ADHD diagnostic system for family dogs. Scientific Reports. https://doi.org/10.1038/s41598-025-09988-8
[5] Barela, Worth & Stevens (2024). Impulsivity as a trait in domestic dogs: A systematic review and meta-analysis. Journal of Comparative Psychology. https://doi.org/10.1037/com0000352
[6] Hecht et al. (2019). Significant Neuroanatomical Variation Among Domestic Dog Breeds. Journal of Neuroscience. https://doi.org/10.1523/JNEUROSCI.0303-19.2019
[7] Ogata et al. (2013). Brain structural abnormalities in Doberman pinschers with canine compulsive disorder. Progress in Neuro-Psychopharmacology & Biological Psychiatry. https://doi.org/10.1016/j.pnpbp.2013.04.002
[8] Tsilioni et al. (2014). Elevated serum neurotensin and CRH levels in children with autistic spectrum disorders and tail-chasing Bull Terriers with a phenotype similar to autism. Translational Psychiatry. https://doi.org/10.1038/tp.2014.106
[9] Cao et al. (2021). Whole genome analyses reveal significant convergence in obsessive-compulsive disorder between humans and dogs. Science Bulletin. https://doi.org/10.1016/j.scib.2020.09.021
[10] Lin et al. (1999). The sleep disorder canine narcolepsy is caused by a mutation in the hypocretin (orexin) receptor 2 gene. Cell. https://doi.org/10.1016/S0092-8674(00)81965-0
[11] Hungs et al. (2001). Identification and functional analysis of mutations in the hypocretin (orexin) genes of narcoleptic canines. Genome Research. https://doi.org/10.1101/gr.gr-1610r
[12] Mondino et al. (2025). Familial Narcolepsy in Dogo Argentino Dogs Is Caused by a Tandem Duplication Mutation in HCRTR2. Journal of Veterinary Internal Medicine. https://doi.org/10.1111/jvim.70056
[13] Morrill et al. (2022). Ancestry-inclusive dog genomics challenges popular breed stereotypes. Science. https://doi.org/10.1126/science.abk0639
[14] Lord et al. (2025). Genetic testing predicts appearance but not behavior in dogs. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2421752122
[15] Freedman, King & Elliot (1961). Critical period in the social development of dogs. Science. https://pubmed.ncbi.nlm.nih.gov/13701603/
[16] McEvoy et al. (2022). Canine Socialisation: A Narrative Systematic Review. Animals. https://doi.org/10.3390/ani12212895
[17] Bouton, Maren & McNally (2021). Behavioral and neurobiological mechanisms of Pavlovian and instrumental extinction learning. Physiological Reviews. https://doi.org/10.1152/physrev.00016.2020
[18] Merck Veterinary Manual. Behavior Problems of Dogs: Hyperkinesis (Hyperactivity) in Dogs. https://www.merckvetmanual.com/behavior/behavior-of-dogs/behavior-problems-of-dogs
[19] Merck Veterinary Manual. Diagnosis of Behavior Problems in Animals. https://www.merckvetmanual.com/behavior/behavioral-medicine-introduction/diagnosis-of-behavior-problems-in-animals