Conflict Between Dogs in the Same Household: What the Evidence Shows

1. The Constellation Nobody Designed

1.1 What Makes a Household Different

Dogs living in the same home share space, routines and access to food, resting places, people and movement through narrow areas. Compared with free-ranging groups, household dogs generally have fewer opportunities to create distance or avoid one another for long periods. That does not mean free-ranging dogs live without stable social structure; observational studies have documented structured relationships in such groups.

That difference matters for how the research transfers. Findings about free-ranging populations describe animals with substantially more opportunities for spatial separation than two dogs sharing a home (how dogs organize themselves when they can choose). Household management therefore has to provide distance deliberately when a pair cannot regulate proximity safely on its own.

1.2 What This Article Covers

Three bodies of evidence exist and they rarely get read together: clinical case series describing dogs referred for fighting with a housemate, owner surveys estimating how common it is in the general population, and observational work on relationships within stable dog groups. Each answers a different question, and the disagreements between them are informative.

1.3 How to Read the Evidence Here

Two limitations run through the field. Clinical samples consist of the cases severe enough to reach a referral clinic, so their proportions describe referral populations rather than dogs in general. And almost everything rests on owner report, which is a problem in a domain where the relevant signals are subtle and the owner is usually part of the situation (how behavior is defined before it is counted).

2. How Common Is It

2.1 The Population Estimate

In a cross-sectional survey of UK dog owners, aggression — defined as barking, lunging, growling or biting — towards other dogs in the same household was reported by 8 percent of owners, against 22 percent towards unfamiliar dogs. Concordance between dog-directed and human-directed aggression was low, indicating that most dogs were not aggressive across contexts (Casey, Loftus, Bolster, Richards & Blackwell, 2013).

That last point deserves emphasis because households routinely generalize in the opposite direction. A dog that fights with its housemate is not thereby a dangerous dog, and a dog that is difficult with strange dogs on walks is not thereby a problem at home.

2.2 What Went With It

In the same survey, aggression towards other household dogs was associated with increasing age of the dog, with the owner's use of positive punishment or negative reinforcement in training, and with attending ring-craft classes. The models explained less than 15 percent of the variance, which the authors take as evidence that most of what separates affected from unaffected households was not measured (Casey et al., 2013).

Both halves belong in any summary. The association with aversive training methods is consistent with the wider literature (what aversive methods do), and the low explained variance means nobody can predict from demographics which household will have this problem.

2.3 The Clinical Picture

At the severe end, a case series from a veterinary behavior clinic recruited 38 pairs of dogs referred specifically for fighting with a housemate. For the 30 cases with injury information available, 15 dogs required veterinary treatment and 3 owners required medical attention after intervening. Fights occurred most often between dogs of the same sex, and the reported aggressor was commonly the younger dog and the more recent arrival (Wrubel, Moon-Fanelli, Maranda & Dodman, 2011).

An earlier series of 99 cases of aggression between dogs described the same overall pattern and remains one of the few sources on treatment and outcome (Sherman, Reisner, Taliaferro & Houpt, 1996).

The largest dataset comes from a record review at a referral behavior clinic covering nine years: 305 pairs, 610 dogs. Seventy percent of the pairs included at least one female, 61.6 percent were same-sex, and resource guarding was the most common fight trigger, at 72.8 percent. Possessive aggression was the most common comorbidity in individual dogs, at 35.4 percent (Feltes, Stull, Herron & Haug, 2020).

2.4 Why the Two Numbers Do Not Contradict Each Other

Eight percent of surveyed owners reporting some household-directed aggression and a referral series containing serious injuries describe different populations, not contradictory facts. The survey counted barking, lunging, growling or biting at any level; referral clinics see cases serious enough for specialist help.

Keeping those populations apart matters in consultation. Occasional growling around a resource and repeated skin-breaking attacks are not interchangeable simply because both meet a broad definition of aggression. Severity, predictability, injury history and the situations in which conflict occurs are more informative than the label alone.

2.5 What the Surveys Cannot Tell Us

Cross-sectional data show associations at one point in time. The survey finding that aversive training methods went with household aggression is compatible with the methods contributing to the problem, with the problem prompting the methods, and with a third factor producing both (Casey et al., 2013).

That is not a reason to ignore it. It is a reason to state it as it stands rather than as a demonstrated cause, and to note that the wider literature on those methods points in the same direction for reasons that do not depend on this survey.

3. Who Fights With Whom

3.1 Sex Combinations

The most consistent finding across clinical series is that same-sex pairs are overrepresented, with female-female pairs particularly affected (Sherman et al., 1996; Wrubel et al., 2011). Observational work points in the same direction from the other side: mixed-sex dyads were more likely to affiliate and less likely to show one-directional submission and aggression than same-sex pairs (Trisko & Smuts, 2015).

The practical reading is narrow. Same-sex pairs are more strongly represented among cases that reach a clinic, which is not the same as saying that a given same-sex pair will have trouble. Most do not.

3.2 Order of Arrival and Age Gap

The clinical series describe the aggressor as typically the younger dog and the later arrival (Wrubel et al., 2011), and the largest dataset quantifies it: the aggressor had been acquired after the recipient in 59.3 percent of 305 pairs, and aggressors were on average 16 months younger and 1.5 kilograms heavier than the dogs they attacked (Feltes et al., 2020).

All three asymmetries matter: in this referral sample the reported aggressor was more often the later acquisition, was younger on average and was slightly heavier on average. None of those variables by itself explains why an individual fight occurred, and the data do not show that a younger newcomer was attempting to "challenge" an older resident.

The practical point is narrower: seniority is not a reliable guide to causation or treatment. The order in which dogs entered the household can be recorded as part of the history, but it should not be converted into a rule that one dog must defer to another (why seniority is not a social rule).

3.3 Size and Injury

Where a size difference exists, the consequences of a fight may be asymmetric. In Hanekom et al. (2025), among 89 dog pairs with size information, the injured dog was larger than the other dog in only 10 percent of pairs and the dogs were similarly sized in 42 percent; in the remaining pairs the injured dog was smaller. The study involved dogs presented for veterinary treatment of bite wounds in Pretoria, South Africa, so the percentages describe that clinical population rather than all multi-dog households.

For practice, size belongs primarily in the injury-risk assessment. A large size mismatch can raise the potential consequence of a single incident even when it does not tell us why the conflict occurs or whether treatment will succeed.

3.4 Neutering

Neutering is the first thing households are told to consider, usually on the assumption that it reduces conflict. The evidence on neutering and behavior generally is mixed and effect-dependent, and it does not support a blanket expectation of improvement in inter-dog conflict (what neutering does and does not change).

The available household-conflict data do not show that neutering reliably resolves an established problem. Hanekom et al. (2025) also found population-specific differences in reproductive status among bite-injury households, including more intact males in affected households, while some common pairings still involved neutered dogs. These descriptive findings do not isolate a causal hormonal mechanism. A decision about neutering therefore belongs with the veterinarian and should not be presented as an evidence-based treatment for an existing household conflict.

4. What the Fights Are About

4.1 Resources, Mostly

The triggers reported in clinical work are unglamorous: food, found items, chews, resting places, narrow passages, human attention or proximity, arrival and departure routines, and periods of high excitement. In the Wrubel follow-up, 14 of 25 responding households identified the presence of a household member as a fight trigger.

That association does not show why the person's presence mattered. Human attention may function as a contested resource in some cases; in others, meals, greetings, leash routines or attempts to intervene may simply occur when people are present. Owner presence is therefore a context to analyze, not a demonstrated cause (how resource guarding is assessed).

4.2 Guarding Between Dogs Is Its Own Category

Resource guarding directed at another dog and resource guarding directed at a person are related but not identical, and the research has separated them. A cross-sectional survey examined factors associated with guarding behavior in the presence of other dogs (Jacobs, Coe, Pearl, Widowski & Niel, 2018a), and an expert-opinion study worked on defining the terms themselves, because "possessive aggression" and "resource guarding" have been used inconsistently (Jacobs, Coe, Widowski, Pearl & Niel, 2018b).

The terminological work matters more than it sounds. If two clinicians mean different things by guarding, the case series they publish cannot be compared, and the household advice that follows from them differs accordingly.

4.3 Owners Miss the Early Part

Jacobs et al. (2017) found that owners were better at recognizing overt aggressive forms of resource guarding than some subtler patterns. Avoidance, rapid consumption and threatening behavior were more difficult to identify consistently, and participants with greater dog knowledge or training experience tended to perform better.

In practice, a report that a fight came "from nowhere" should prompt closer reconstruction rather than an assumption about what happened beforehand. A longer sequence of subtle behavior may have been missed, but rapid escalation or other mechanisms are also possible (what the visible signals do and do not show).

4.4 Arousal and Redirection

Clinical histories also describe fights around highly arousing events such as arrivals, departures or other excitement. High arousal can therefore be an important context, but the available household studies do not quantify how often true redirected aggression accounts for these incidents.

The distinction matters because an external trigger and conflict over a resource are not automatically the same mechanism. When fights cluster around windows, greetings or leash routines, the external event, spatial arrangement and each dog's response should be analyzed separately rather than labeled "redirection" by default (why arousal is a variable in its own right).

4.5 The Same Trigger Does Not Mean the Same Problem

Two households can report fights at the food bowl and need different plans: one where a dog guards the bowl and the other approaches, and one where the dogs collide in a doorway on their way to it. The trigger list is a starting point for observation rather than a diagnosis, which is why the recording described later matters more than the label attached to the case.

5. Do Household Dogs Form a Hierarchy?

5.1 What Has Been Observed

In a year-long observational study of a group of domestic dogs, many dyads did not provide enough directional interactions to classify a relationship confidently, but the relationships that could be assessed were largely unidirectional and the group showed significant linearity. Rank correlated with age, with older dogs tending to rank higher, and not with size. Mounting and chin-over behaviors were not reliable status indicators, while muzzle licking was consistently directed from lower- to higher-ranking individuals (Trisko & Smuts, 2015).

Free-ranging groups can also show structured dominance relationships. Cafazzo et al. (2010) described a significantly linear hierarchy in a group of free-ranging dogs in Rome, related to age, sex and competitive context. These studies support the existence of dominance relationships in at least some dog groups; they do not show that every household dyad has a clear hierarchy or that hierarchy explains an individual fight.

5.2 What Owners Perceive

Owners of multi-dog households readily assign status labels, and what those labels track has been examined directly. In a questionnaire study, the personality factor labeled assertiveness accounted for 34 percent of the variance in owner-assigned dominance status, trainability for 5 percent and age for 4 percent. Dogs rated as dominant did not score higher on intraspecific aggression (Wallis, Iotchev & Kubinyi, 2020).

A related study reported that owner ratings of dominance corresponded to ethologically defined markers (Kubinyi & Wallis, 2019). Taken together, owner labels are not pure invention — they track something observable — while also loading heavily on a personality impression rather than on a relationship between two specific animals.

5.3 Why This Does Not License the Usual Advice

The classic recommendation derived from hierarchy thinking is to support the dog labeled "dominant": feed it first, greet it first or systematically give it priority. Existing studies do not establish this as an effective treatment.

Dominance is a relationship between individuals, not a diagnosis for why a particular fight occurred. Owner-assigned status is influenced strongly by perceived assertiveness (Wallis et al., 2020), although other work shows that owner judgments can also correspond to ethologically relevant markers such as resource access and fight outcomes (Kubinyi & Wallis, 2019). Wrubel et al. (2011) frequently included prioritized access for one dog in a multi-component treatment package, but because treatments were combined and there was no control group, the contribution of that component cannot be isolated. The evidence therefore neither validates rank-based treatment nor requires denying that dominance relationships can exist.

5.4 What Free-Ranging Groups Do and Households Cannot

Free-ranging and daycare groups generally offer more opportunity for spatial separation than a household pair confined to the same rooms. The Rome study describes structured relationships in animals moving over a larger area (Cafazzo et al., 2010), and dogs in the daycare study could disengage and move away (Trisko & Smuts, 2015).

That difference limits direct transfer. Social-structure findings from those settings can inform how relationships are described, but household management still has to account for the much tighter control of distance and access. Providing safe opportunities for separation is therefore a practical environmental intervention, not an inference that hierarchy disappears in homes.

5.5 Why the Label Matters Less Than the Contingency

Whether or not a particular pair has a dominance relationship, what changes their behavior day to day is which situations reliably produce which outcome. A dog that has learned that approaching the other at the bowl leads to a fight behaves differently from one that has learned it leads to being given space.

That is why the practical plan in Section 8 is built around arrangements rather than around status: contingencies can be changed by the household, a social label cannot (how contingencies shape behavior).

6. The Medical Layer

6.1 Pain Changes the Group

Conflict that appears or worsens without an obvious environmental change should increase the priority of medical assessment. Pain can alter tolerance of touch, movement, proximity and interruption, and can therefore contribute to aggressive behavior in some dogs (how chronic pain changes behavior). Direct studies quantifying pain specifically in intrahousehold inter-dog aggression are still lacking.

The pattern to look for is a change in an established pair rather than a long-running problem: dogs that lived together peacefully for years and started fighting, or a dog that suddenly objects to being approached at rest.

6.2 Age and Cognitive Change

Older dogs figure in this literature twice: age was associated with household-directed aggression in the population survey (Casey et al., 2013), and older dogs tend to be rated and to rank higher in the group work (Trisko & Smuts, 2015; Wallis et al., 2020). An aging dog that becomes less able to read signals or to move out of the way changes the group's dynamics without doing anything aggressive itself (what changes in aging dogs).

6.3 Demographic and Environmental Factors

Large questionnaire studies of purebred dogs have examined how demographic, environmental and behavioral factors relate to aggressive behavior generally (Mikkola et al., 2021). Such work is useful for describing distributions across a population, and it does not identify which dog in which household will have trouble with its housemate.

6.4 What to Rule Out First

A veterinary examination of both dogs belongs early in any serious or newly changed case, rather than only after behavior work has failed. Orthopedic pain, dental pain, ear disease, gastrointestinal discomfort, neurological change and sensory decline can alter responses to approach or contact even when the fight itself does not reveal the medical contribution.

The practical implication is not that every household conflict is medical. It is that untreated pain or disease can undermine a behavior plan and should remain part of the differential assessment while safety management proceeds (how pain changes responses to handling and proximity).

7. What the Evidence Says About Treatment

7.1 The Honest Summary

There is no controlled trial establishing an effective treatment for aggression between household dogs. What exists is case-series outcome data. In the 38-pair series, treatment was typically multi-component: prioritized resource access for one dog was recommended in 95 percent of cases, a "Nothing in Life Is Free" program in 84 percent, headcollar or trailing-leash management in 79 percent, psychoactive medication in 79 percent, obedience training in 71 percent and trigger avoidance in 61 percent. Only 25 of 38 households returned the follow-up questionnaire, and owners reported a median improvement in the overall problem of 69 percent (Wrubel et al., 2011).

Those data show that improvement can occur in referred cases, but they cannot identify which component caused it. There was no untreated comparison group, multiple interventions were used simultaneously, and follow-up was incomplete. The 99-case series likewise provides clinical outcome information rather than controlled efficacy evidence (Sherman et al., 1996; how behavioral protocols are tested).

7.2 What Was Associated With a Poor Outcome

The largest dataset examined this directly, with a poor outcome defined as permanent separation, rehoming or euthanasia. Three factors were associated with it: same-sex pairs, a history of bites that broke the skin, and aggression on sight of the other dog rather than in a contested situation. The authors conclude that such pairs need behavioral intervention quickly because of the high risk of a poor outcome (Feltes et al., 2020).

To that list belongs a practical fourth: the household's ability to keep the dogs apart safely while anything is being trained.

A household in which one dog cannot enter a room without an immediate attack is in a different situation from one with occasional disputes at the food bowl, and treating them with the same plan is the most common error in this area.

7.3 Medication

Psychoactive medication appears in the clinical recommendations from these series, and the general canine literature on medication for behavior problems is considerably better developed for separation-related problems and noise aversion than for inter-dog conflict (what the canine drug trials show). Nothing in the existing evidence supports a specific drug for fighting between housemates, and the decision belongs to a veterinarian who has examined both dogs.

7.4 Outcomes Nobody Likes to Name

The outcomes in this problem area are not only better or worse behavior. They include permanent separation within the home, rehoming one dog, and euthanasia, and households facing repeated injuries are weighing those options whether or not anyone says so out loud.

A consultation that does not name them leaves the household to decide alone and usually later than necessary. Saying early which severity markers make permanent separation or rehoming a realistic outcome is not defeatism; it is what allows a household to plan rather than to react after the next incident.

7.5 Why Prognosis Talk Has to Be Careful

Without controlled trials there is no basis for a general success percentage. Feltes et al. (2020) found three factors associated with the defined poor outcome: same-sex pairing, a history of skin-breaking bites and aggression on sight. Size mismatch belongs in the injury-risk assessment, but it was not established there as a prognostic factor for rehoming, euthanasia or permanent separation. Likewise, a long history of fighting may matter clinically but should not be presented as one of the study's validated outcome factors.

Improvement in the available case-series literature was measured mainly as fewer or less severe fights, not as restoration of an affiliative relationship. A serious-case plan can therefore aim for safe coexistence or durable separation without assuming that the dogs must become friends (Wrubel et al., 2011).

That last distinction is worth making explicit in a household that expects the dogs to become friends. The realistic target in a serious case is a household that functions safely, not a restored relationship.

8. What Follows for Practice

8.1 Management Comes First and Is Not a Failure

Physical management — for example gates, doors, rotation or secure separate areas — is the most reliable way to prevent direct access while a serious case is assessed. It is not a treatment failure; it creates the conditions in which medical work-up, observation and controlled training can occur without repeated injuries. Training can proceed with dogs separated or under carefully controlled exposure, but repeated uncontrolled fights should not be part of the process (why rehearsed behavior is hard to remove).

8.2 Predictability Beats Fairness

Households often focus on equal treatment, but the clinical trigger lists suggest a more useful target: reduce simultaneous competition and make high-risk routines easy to predict. Feed dogs in separate locations, distribute chews only where they cannot be contested, and stagger movement through narrow doors or corridors when necessary.

This is a clinical management principle rather than a demonstrated treatment effect. Consistency can reduce opportunities for conflict, but the evidence does not show that a fixed priority order — such as always sending one particular dog through a doorway first — treats the underlying problem.

8.3 Rearrange the Places, Not the Relationship

Most of the trigger list is geography. Doorways, corridors, stairs, the space in front of the fridge and the two-meter radius around the person are where conflicts happen because they force proximity. Feeding in separate rooms, adding a second exit to a resting area, widening a route and removing chews from shared space address the situation rather than the dogs.

8.4 Watch the Moments Before

Because subtle resource-related behavior can be harder for owners to identify (Jacobs et al., 2017), a useful record focuses on what preceded an incident: where each dog was, what was present, who entered or moved, who had access to which resource, and what each dog did immediately beforehand. A structured log can reveal recurring configurations without assuming in advance that every conflict has the same cause.

8.5 What Not to Do

Several common responses deserve caution. Punishing growling can suppress warning behavior or add an aversive consequence without resolving the underlying motivation (what punishment costs). Deliberately allowing repeated fights in the hope that the dogs will "sort it out" exposes both animals to preventable injury. And systematically supporting whichever dog has been labeled "dominant" is not an evidence-based standalone treatment, even though dominance relationships can exist between dogs.

8.6 When to Refer

Skin-breaking bites, attacks on sight, injuries to people intervening, repeated escalation or an inability to maintain safe separation are strong reasons for early specialist referral. In Wrubel et al. (2011), 15 of 30 cases with injury data required veterinary treatment for dogs and 3 of 30 involved owners needing medical attention. In Feltes et al. (2020), skin-breaking bites and aggression on sight were among the factors associated with permanent separation, rehoming or euthanasia.

8.7 Adding a Second Dog

The prevention question comes up constantly and has thinner evidence than the treatment question. What exists is the composition of clinical samples: same-sex pairs and particularly female-female pairs are overrepresented there, as are pairs where the younger dog arrived later (Sherman et al., 1996; Wrubel et al., 2011). No prospective study has followed newly formed pairs to see which combinations actually develop problems.

The authors of the largest referral series advised against same-sex pairings in pre-adoption counseling (Feltes et al., 2020), but that recommendation is stronger than the retrospective design can establish. No prospective study has shown that a particular combination of sex and age reliably prevents conflict. Clinical sample composition can inform a risk discussion, not provide a guarantee.

The resident dog's current behavior and health still matter. Existing dog-directed resource guarding, significant pain, severe reactivity or difficulty coping with proximity are reasons for careful assessment before adding another dog. They do not prove that successful multi-dog living is impossible, but they can increase management demands and should not be expected to improve simply through exposure to a permanent housemate (how guarding is assessed beforehand).

8.8 What to Tell a Household

Three sentences cover most of it. Fighting between housemates is a common problem and not a verdict on either dog or on the household. Separating the dogs while a plan is built is the first step rather than the last resort. And what will be measured is whether incidents get rarer and milder, not whether the dogs end up liking each other.

9. Four Household Situations

9.1 The New Dog and the Resident

Typical picture: a young dog joins a household with an older resident, things go well for weeks or months, and conflict appears as the newcomer matures. The clinical pattern — aggressor younger and more recently acquired (Wrubel et al., 2011) — fits this case, and the household's expectation that the resident will "put the youngster in its place" usually does not.

The practical approach is to reduce contested moments before they are needed rather than after the first fight, and to resist the temptation to referee.

9.2 The Pair That Lived Together for Years

Typical picture: two dogs with a long peaceful history begin fighting. A new change in an established pair should raise the priority of medical assessment in either dog — including pain, sensory decline or cognitive change — without assuming that a medical cause has already been established (how internal pain changes behavior).

9.3 The Mismatched Pair

Typical picture: a large and a small dog, or a mature and a very young one, with conflict that looks one-sided. Here the injury risk dominates the plan regardless of how often the disputes occur, and management is not negotiable while anything else is being worked on.

9.4 The Pair That Only Fights in the Owner's Presence

Typical picture: the dogs are reported to be calm when alone and to fight when people are home. That pattern appears in clinical reports, but it does not show that the dogs are performing for an audience or that the person is necessarily the contested resource.

Possible explanations include competition over human attention, routines that occur only when people are present, intervention by household members, or arousal around greetings, food or walking. The useful task is to identify which human-linked situations precede conflict rather than to infer motive from owner presence alone.

9.5 Three Dogs or More

Typical picture: a household with several dogs where two animals clash and another dog also becomes involved. The current clinical literature is built largely around pairs, so mechanisms of third-dog involvement have not been established well enough to claim that a dog "joins" because of noise, arousal or social alignment.

For practice, multi-dog incidents justify conservative management: separation and reintroduction plans should consider every dog that had access to the event, not only the original pair, until the recurring configurations are understood.

9.6 What to Record in All of These

The same short log works across situations: date and time, where each dog was, what was present, who else was in the room, what happened immediately before the incident, how it ended and how long each dog took to settle. Repeated records can reveal recurring configurations; the evidence does not support a fixed number of days or a fixed number of patterns that should emerge.

10. Summary at a Glance

About 8 percent of owners report aggression towards a household dog — Against 22 percent towards unfamiliar dogs, with low concordance between dog-directed and human-directed aggression (Casey et al., 2013).

Known factors explain little — The models in that survey accounted for under 15 percent of the variance, meaning most of what distinguishes affected households was not measured (Casey et al., 2013).

Aversive training methods were associated with household aggression — Along with increasing age of the dog (Casey et al., 2013).

Referral cases can be severe — Among the 30 Wrubel cases with injury data, 15 involved veterinary treatment for dogs and 3 involved owners needing medical attention after intervening (Wrubel et al., 2011).

Same-sex pairs dominate the clinical picture — 61.6 percent of 305 referral pairs were same-sex and 70 percent included at least one female (Feltes et al., 2020; see also Sherman et al., 1996; Wrubel et al., 2011).

Resource guarding is the most common trigger — In 72.8 percent of those pairs, with possessive aggression the most common comorbidity at 35.4 percent of individual dogs (Feltes et al., 2020).

Three features were associated with a poor outcome — Same-sex pairing, skin-breaking bites and aggression on sight, where poor outcome meant permanent separation, rehoming or euthanasia (Feltes et al., 2020).

The aggressor is often the newer, younger dog — Acquired after the recipient in 59.3 percent of pairs, on average 16 months younger and 1.5 kilograms heavier (Feltes et al., 2020).

Owner presence can be part of the trigger context — In the Wrubel follow-up, 14 of 25 responding households identified a household member's presence as a fight trigger; the study did not establish why that association occurred.

Owners recognize overt guarding better than subtler patterns — Avoidance, rapid consumption and threatening behavior were harder to identify consistently (Jacobs et al., 2017).

Dominance relationships can exist between dogs — Observational studies report directional and sometimes linear structures, but that does not establish the cause of a household fight or validate rank-based treatment (Cafazzo et al., 2010; Trisko & Smuts, 2015).

Owner-assigned status reflects both perception and behavior — Assertiveness accounted for 34 percent of the variance in one study, while another found correspondence with ethologically relevant markers such as resource access and fight outcomes (Wallis et al., 2020; Kubinyi & Wallis, 2019).

Size belongs in the injury-risk assessment — In the Pretoria bite-injury sample, the injured dog was larger than its opponent in only 10 percent of 89 pairs with size information; these figures are population-specific (Hanekom et al., 2025).

Guarding between dogs is a defined field of its own — With work on associated factors and on the terminology itself, because "possessive aggression" and "resource guarding" have been used inconsistently (Jacobs et al., 2018a, 2018b).

No treatment has been tested against a control — Outcome data come from case series, in a problem where households weigh rehoming and euthanasia (Sherman et al., 1996; Wrubel et al., 2011).

11. Research Gaps and Critical Appraisal

No controlled treatment trial exists. Everything known about outcomes comes from case series without control groups (Sherman et al., 1996; Wrubel et al., 2011), in a problem area where households face decisions about rehoming and euthanasia.

The largest dataset is a record review. The 305-pair series extracted data retrospectively from clinic records over nine years (Feltes et al., 2020), which constrains what was recorded and how consistently, and it reports associations with outcome rather than effects of treatment.

Clinical samples are not the population. Referral series describe the severe end; the survey estimate of 8 percent describes a different population, and the two are routinely quoted as if they described the same thing.

Much of the literature depends on owner report. Jacobs et al. (2017) showed that recognition differs by behavioral pattern: overt aggression was easier to identify than subtler forms such as avoidance or rapid consumption. That makes owner report useful but imperfect rather than uniformly unreliable.

The direction of the training association is unknown. Aversive methods were associated with household aggression (Casey et al., 2013), and a cross-sectional survey cannot say whether the methods contributed to the aggression or the aggression prompted the methods.

Household group structure is barely studied. Key observational work on dominance relationships used daycare and free-ranging groups rather than the two-, three- and four-dog households where clinical conflict occurs (Trisko & Smuts, 2015; Cafazzo et al., 2010). Third-dog involvement in household fights remains particularly poorly described.

Prevention is untested. Advice on matching a second dog by sex, age, size or temperament rests largely on the composition of clinical samples rather than prospective studies following newly formed pairs.

The role of the owner is undermeasured. Owner presence appears as a trigger in case series, but no study has experimentally varied owner behavior in the household and measured the effect on conflict between the dogs.

Risk patterns may not transfer unchanged between populations. Hanekom et al. (2025) studied bite injuries presented to a veterinary teaching hospital in Pretoria and reported some patterns that differed from earlier Western referral samples. Population, household structure, neuter status, breed distribution and access to veterinary care can all affect which cases enter a dataset.

Injury risk and treatment prognosis are different questions. Size mismatch plausibly changes the potential severity of an incident and appears relevant in bite-injury data, but it has not been validated as a predictor of treatment outcome in the way that same-sex pairing, skin-breaking bites and aggression on sight were associated with poor outcome in Feltes et al. (2020).

12. Conclusion

Conflict between dogs in the same household ranges from mild resource-related signalling to serious repeated injury, and the available studies describe those different parts of the spectrum with very different samples. In a UK owner survey, 8 percent reported some aggression towards a household dog, but the models explained less than 15 percent of the variance (Casey et al., 2013). At the referral end, Wrubel et al. (2011) reported veterinary treatment for dogs in 15 of 30 cases with injury data and medical treatment for owners in 3 of 30. In the largest referral record review — 305 pairs over nine years — 61.6 percent were same-sex, resource guarding was recorded as a trigger in 72.8 percent, the aggressor had arrived later in 59.3 percent, and same-sex pairing, skin-breaking bites and aggression on sight were associated with permanent separation, rehoming or euthanasia (Feltes et al., 2020).

The social-structure literature does not support either extreme. Dominance relationships can be measurable in dog groups and some dyads, but the existence of a dominance relationship does not explain why a specific fight happened and does not validate rank-based treatment. Likewise, owner judgments are not pure invention, but they are influenced strongly by traits such as assertiveness (Cafazzo et al., 2010; Kubinyi & Wallis, 2019; Trisko & Smuts, 2015; Wallis et al., 2020).

What follows for practice is therefore deliberately conservative: prevent uncontrolled access when injury risk is meaningful, analyze recurring situations rather than infer motives from labels, reduce competition and forced proximity, include medical assessment when behavior changes, and refer early when bites break skin or attacks occur on sight. Size mismatch should increase attention to injury risk, while population-specific findings such as Hanekom et al. (2025) should not automatically be exported to every country or household. What the field still cannot offer is a controlled treatment trial establishing which intervention package works best.

Key Insights (Takeaways)

• Household conflict is not simply leash reactivity at home: aggression towards housemates and unfamiliar dogs showed low concordance in the UK survey (Casey et al., 2013).

• In referral data, aggressors were often younger and later acquisitions, but those features are associations and do not reveal a motive such as "challenging" the resident dog (Wrubel et al., 2011; Feltes et al., 2020).

• Human presence can be part of the trigger context, but the evidence does not show that the person is necessarily the resource or the cause of the fight.

• Owners identify overt aggressive guarding better than some subtler resource-related behaviors, so structured observation is more useful than assuming every "sudden" fight had a hidden warning sequence (Jacobs et al., 2017).

• Dominance relationships can exist between dogs; they do not by themselves diagnose the cause of a fight or justify feeding, greeting or moving one dog first as treatment.

• Size mismatch belongs in injury-risk assessment. Hanekom et al. (2025) adds useful bite-injury data, but its Pretoria sample should not be treated as a universal household-risk profile.

• Same-sex pairing, skin-breaking bites and aggression on sight were associated with poor outcome in Feltes et al. (2020); they were not a validated prediction formula.

• Physical management is often the safest first step while assessment and controlled training proceed. No controlled trial has yet established one treatment package as effective for this problem.

References

Cafazzo, S., Valsecchi, P., Bonanni, R., & Natoli, E. (2010). Dominance in relation to age, sex, and competitive contexts in a group of free-ranging domestic dogs. Behavioral Ecology, 21(3), 443–455. https://doi.org/10.1093/beheco/arq001

Casey, R. A., Loftus, B., Bolster, C., Richards, G. J., & Blackwell, E. J. (2013). Inter-dog aggression in a UK owner survey: Prevalence, co-occurrence in different contexts and risk factors. Veterinary Record, 172(5), 127. https://doi.org/10.1136/vr.100997

Feltes, E. S. M., Stull, J. W., Herron, M. E., & Haug, L. I. (2020). Characteristics of intrahousehold interdog aggression and dog and pair factors associated with a poor outcome. Journal of the American Veterinary Medical Association, 256(3), 349–361. https://doi.org/10.2460/javma.256.3.349

Hanekom, J., Thompson, P. N., Schoeman, J. P., & Van Vollenhoven, E. (2025). Household hostilities: A descriptive study of inter-dog aggression requiring veterinary treatment of dog bite wounds in Pretoria, South Africa. Applied Animal Behaviour Science, 292, 106819. https://doi.org/10.1016/j.applanim.2025.106819

Jacobs, J. A., Coe, J. B., Pearl, D. L., Widowski, T. M., & Niel, L. (2018a). Factors associated with canine resource guarding behaviour in the presence of dogs: A cross-sectional survey of dog owners. Preventive Veterinary Medicine, 161, 134–142. https://doi.org/10.1016/j.prevetmed.2017.02.004

Jacobs, J. A., Coe, J. B., Widowski, T. M., Pearl, D. L., & Niel, L. (2018b). Defining and clarifying the terms canine possessive aggression and resource guarding: A study of expert opinion. Frontiers in Veterinary Science, 5, 115. https://doi.org/10.3389/fvets.2018.00115

Jacobs, J. A., Pearl, D. L., Coe, J. B., Widowski, T. M., & Niel, L. (2017). Ability of owners to identify resource guarding behaviour in the domestic dog. Applied Animal Behaviour Science, 188, 77–83. https://doi.org/10.1016/j.applanim.2016.12.012

Kubinyi, E., & Wallis, L. J. (2019). Dominance in dogs as rated by owners corresponds to ethologically valid markers of dominance. PeerJ, 7, e6838. https://doi.org/10.7717/peerj.6838

Mikkola, S., Salonen, M., Puurunen, J., Hakanen, E., Sulkama, S., Araujo, C., & Lohi, H. (2021). Aggressive behaviour is affected by demographic, environmental and behavioural factors in purebred dogs. Scientific Reports, 11, 9433. https://doi.org/10.1038/s41598-021-88793-5

Sherman, C. K., Reisner, I. R., Taliaferro, L. A., & Houpt, K. A. (1996). Characteristics, treatment, and outcome of 99 cases of aggression between dogs. Applied Animal Behaviour Science, 47(1–2), 91–108. https://doi.org/10.1016/0168-1591(95)01013-0

Trisko, R. K., & Smuts, B. B. (2015). Dominance relationships in a group of domestic dogs (Canis lupus familiaris). Behaviour, 152(5), 677–704. https://doi.org/10.1163/1568539X-00003249

Wallis, L. J., Iotchev, I. B., & Kubinyi, E. (2020). Assertive, trainable and older dogs are perceived as more dominant in multi-dog households. PLoS ONE, 15(1), e0227253. https://doi.org/10.1371/journal.pone.0227253

Wrubel, K. M., Moon-Fanelli, A. A., Maranda, L. S., & Dodman, N. H. (2011). Interdog household aggression: 38 cases (2006–2007). Journal of the American Veterinary Medical Association, 238(6), 731–740. https://doi.org/10.2460/javma.238.6.731