Introducing a cat to a dog: what the evidence shows, and what to do in an Indian flat

Does introducing a cat to a dog work? Three surveys, three denominators

Introducing a cat to a dog ends amicably in roughly two thirds of households, and the three published owner surveys that counted it agree with each other far more than the stereotype does.

The oldest of the three is the most direct. Feuerstein and Terkel collected 170 questionnaires from households keeping both species in Israel and sorted each relationship into one of three categories. 66.5 percent of the dogs and 65.5 percent of the cats were amicable towards the resident animal of the other species, 24.5 percent of dogs and 26 percent of cats were indifferent, and 9 percent of dogs and 8.5 percent of cats were aggressive.

A larger online survey from the University of Lincoln reached 748 mixed-species homes. Most owners believed their cat and dog were comfortable in each other's presence and showed amicability, and the regression turned up something no step list mentions: perceptions of amicability were influenced more by cat factors than by dog factors, and the cat appeared to be the main controller in determining amicability.

The largest is Italian. 1,270 people who owned at least one dog and at least one cat answered a questionnaire between May and July 2014. 62.4 percent of the pairs played together, 68.5 percent slept near each other at least occasionally, and 42.8 percent of the dogs licked the cat against 34.3 percent of cats licking the dog.

Amicable is a category, not a friendship

Indifference is the second most common outcome, at 24.5 percent of dogs and 26 percent of cats in the only study that measured all three categories. Two animals that use the same flat and ignore each other have succeeded. Nothing on this page is aimed at producing the photograph of a cat asleep on a labrador, and treating that as the target is how people keep pushing an introduction that was already finished.

All three surveys are owner reports rather than direct observation, and the Italian authors say so themselves: the questionnaire inevitably involves a degree of subjectivity and some biases in the owners' responses, and voluntary participation plus an imbalance in the gender of participants, 91.2 percent of whom were women, could aggravate those biases. Treat two thirds as the shape of the answer rather than a precise probability for your household.

The two predictors nobody reports: which animal came first, and how old they were

The order of acquisition and the age at the first meeting are the two measured predictors of whether a cat and a dog end up amicable, and almost no page explaining how to introduce them mentions either.

Feuerstein and Terkel state both in one sentence of their own abstract. The animals' gender had little influence on the nature of their interrelationship, while adoption of the cat prior to the dog appears to conduce to establishing an amicable relationship, as does their first encounter taking place at an early age, up to 6 months of age in cats and up to 1 year in dogs. The same study found that the earlier the age of first encounter, the better each animal understood body language that means the opposite thing in the other species.

The Lincoln survey found the age effect independently. Among the variables that survived into its final model were the age at which the cat was introduced to the dog, with a younger age predicting greater perceived amicability, and the comfortability of the cat, which was a stronger predictor than the comfortability of the dog.

The third dataset is the only one that puts a number on the thing you can still change. Generation Pup, a Dogs Trust cohort, had 4,678 valid responses, of which 26.7 percent, 1,248 puppies, lived with at least one cat, and 1,211 of those had already been introduced. Three variables survived into the final multivariable model of 1,201 puppies: each additional dog in the household raised the odds of only desirable behaviour towards the cat to 1.31, 95 percent CI 1.16 to 1.46; a gradual or quite gradual introduction raised them to 1.86, CI 1.14 to 3.05, P equal to 0.013; and being under 12 weeks old when the data were collected raised them to 2.52, CI 1.60 to 3.96.

What to do when the order is already wrong

Most Indian households reading this already have an adult cat and are bringing in a dog, or the reverse, and both of the age predictors have closed. That is worth knowing rather than worth despairing over, because the one factor the Dogs Trust model found that you still own is the speed of the introduction, and it is the factor with the clean confidence interval.

1. Count the weeks you have before the second animal arrives and spend them setting the flat up, not reading more step lists.
2. Pick the slower branch at every decision, because the measured benefit attaches to gradual rather than to any particular technique.
3. Treat the cat's comfort as the outcome you are managing, since cat factors outweighed dog factors in the only regression that tested both.
4. Accept indifference as a finished result rather than a stalled one.
5. Write down the date you started, because three weeks in, nobody remembers whether things are better or just familiar.

The staged protocol: separation, scent swapping, sight through a barrier, shared space

To run a staged introduction, separate the two animals completely, exchange scent before either sees the other, allow sight through a barrier neither can cross, and only then share a room with the dog on a loose lead and the cat free to leave.

The ladder itself is not where this page adds anything. Every shelter page carries a version of it, and our guide to bringing a kitten home in India sets out International Cat Care's four steps in full for the case of a new kitten arriving to a resident dog. What no step list carries is the measurement of what happens when people skip it.

The Dogs Trust cohort asked owners how fast they had gone and recorded four answers. 243 of 1,211, that is 20.1 percent, put the animals together straight away. 229, 18.9 percent, controlled the interaction for up to the first 2 hours. 252, 20.8 percent, went quite gradually over the first day. 487, 40.2 percent, introduced them slowly over a period of more than one day. Nearly two in five households therefore had a loose puppy and a resident cat in the same space inside two hours, and that choice carried an odds ratio.

1. Separate completely. The new animal gets one closed room with every resource it needs duplicated, and the resident animal keeps its own territory unchanged. Nothing is shared and nothing is moved. Expect this to run 3 to 7 days.
2. Swap scent before anyone sees anyone. Move a slept-on cloth from each animal's bed into the other's space daily, and let each investigate at its own pace. Scent swapping is cheap, it is the one stage that costs you nothing, and it is the stage people cut. Expect 3 to 7 days, overlapping stage one.
3. Open the space in turn. Shut the dog away and let the cat walk the whole flat, then reverse it. The cat needs to map the dog's smell onto rooms before it maps it onto a dog. Expect 2 to 5 days.
4. Allow sight through a barrier. A baby gate, a stair gate or a closed grille door lets both animals see each other with no possibility of contact. Keep each session to a few minutes, with both animals occupied and the dog fed or worked away from the gate, and end while both are still calm. Expect 1 to 3 weeks of short daily sessions.
5. Share a room. The dog wears a harness and a loose lead held by a second person, the cat has an open door to retreat through and a high perch to reach, and nobody carries anybody. Build from minutes to an hour over weeks, and keep supervision absolute until the dog ignores the cat without being asked. Expect 2 to 6 weeks.
6. Leave them alone together only when the dog has shown no chasing, no staring and no fixation across many supervised hours, and the cat is eating, using the tray and sleeping on its normal schedule with the dog in the room.

What ends a session early, every time

Four signals end a session before it can escalate, with no discussion. A tight lead means the dog is already over threshold. A hard stare that survives its own name being called is the behaviour that precedes a chase. A cat that bolts has just taught the dog that running happens here. And a cat that stops eating, stops using the litter tray or stops coming out at all has given you a medical and behavioural signal at once, and the Dogs Trust authors note that owners routinely accept exactly these signs as normal when they indicate stress in cats.

The asymmetry: the dog can kill the cat, and the cat can blind the dog

The physical risk runs in both directions and it is not symmetrical, because the cat is the animal that can be killed and the dog is the animal that can lose an eye.

Owners of 748 mixed-species homes reported both directions. 56.5 percent of cats had threatened the dog and 9.6 percent had caused it injury. 18 percent of dogs had threatened the cat and 0.9 percent had caused injury. Asked how often each animal looked uncomfortable in the other's presence, 20.5 percent of cats and 7.3 percent of dogs were uncomfortable every day to once a week.

Read those two injury rates the right way round. A cat injures the dog about ten times as often as the dog injures the cat, and the dog is still the dangerous animal, because the two injuries are not the same injury. A cat's correction is a swipe that lands and stops. A dog's grab is a grip on a four-kilogram animal.

What sets it off is almost always movement. In the Italian survey, asked what their dog did at the sight of a cat, owners of 1,270 dogs reported attacking in 37.2 percent, barking in 23.8 percent, tail-wagging in 20.5 percent and growling in 12.8 percent, with exactly one dog ignoring the cat. Asked the reverse, owners of 1,245 cats reported escaping in 31.3 percent, ignoring in 24.5 percent, hissing in 21.4 percent and approaching amicably in 13.5 percent, with 1.7 percent attacking. A cat that escapes is a cat that runs, and running is the single most reliable trigger for chasing.

The pheromone trial run at Lincoln recruited homes where owners saw room for improvement and found the same pattern in its own baseline: over half of the relationships could be described as the dog being interested and the cat being fearful.

Why the puppy is the one that loses an eye

A retrospective hospital series of ocular cat claw injuries in dogs found very young dogs over-represented, and gave the mechanical reason. Puppies up to the age of 12 weeks have no menace response, so an important protective mechanism is missing in early life, and older dogs are less likely to sustain such an injury. The injuries themselves are corneal and lens capsule perforation with ensuing severe phacoclastic uveitis, and in that series phthisis bulbi, meaning a shrunken blind globe, or secondary glaucoma was the outcome in 50 percent of cases.

Two later series put the same injury in a wider frame. A review of 33 eyes treated by phacoemulsification after corneal and lens laceration between 2000 and 2010 found 27 of them canine, with cat scratch trauma the identified cause in 9 of 33 and the affected dogs younger, at a median 18 months, than the affected cats at 30 months; all the cats and 85.2 percent of the dogs were visual at the last examination. A larger retrospective of 77 patients with the same injury, 67 of them dogs, concluded that the length of the corneal laceration and the time interval from injury to referral were both relevant prognostic factors.

The mitigations differ because the injuries differ. Protecting the cat means removing the chase: distance, a barrier, a loose lead and an exit. Protecting the dog means keeping its face out of claw range, which in practice means never letting a puppy push its nose into a cornered cat, and keeping the cat's claws trimmed at the tips rather than removed. Declawing is covered, with the veterinary position on it, in our guide to making a house cat friendly.

Some dogs cannot safely live with a cat, and here is how you would know

Some dogs cannot safely live with a cat, and the reason is that predatory behaviour arrives without the warning that every other kind of canine aggression gives you.

The ASPCA's classification is blunt about it. Pet dogs still show some classic canine predatory behaviours, including chasing and grabbing fast-moving things; they might chase pets, wildlife and livestock; and some dogs bite and even kill if they manage to catch the thing they are chasing. Predatory aggression is very different from other classifications of aggression because there is rarely any warning before an attack, and a predatory dog does not growl or show her teeth first to warn her victim, so it can seem to come out of the blue.

That is the whole problem with the common advice to watch for growling. A dog that growls at the cat is telling you something, and a dog that goes silent and still while it tracks the cat across a room is telling you something worse. Among the five risk factors the ASPCA lists for deciding whether to live with an aggressive dog, predictability is the one that matters most here, because dogs at the highest risk of being euthanised for aggression are those who give little or no warning before they bite and who are inconsistently, unpredictably aggressive. The other four are size, age, bite history and severity. A dog whose reaction to the cat is predictable, even when the reaction is a bad one, is easier to manage than a dog that is fine for a fortnight and then is not.

How common is it? The four datasets on this page disagree in a useful way rather than a confusing one. 9 percent of dogs were classed as aggressive towards the resident cat in the Israeli survey. 18 percent of dogs had threatened the cat and 0.9 percent had injured it in the UK survey. 1.7 percent of puppies were reported aggressive towards the resident cat, although 48.6 percent chased it and 56.6 percent were overexuberant with it. And 37.2 percent of Italian owners said their dog attacked at the sight of a cat. Hostility in a settled household sits in the single digits. Chasing is close to universal in puppies. Those are two different facts and they are usually reported as one.

Five signals say stop and get professional help rather than carry on with the protocol.

• Silent, still fixation on the cat, with no growl, that does not break when the dog's name is called.
• A chase that continues after the cat has stopped running and is cornered or under furniture.
• A grab, a shake, or any contact that leaves a puncture, however small.
• The same intensity on day 20 as on day 1, with no change under a gradual programme.
• Fixation that transfers outdoors, because a dog can be reliable with the cat indoors and hunt it on a terrace.

Where that is the picture, the honest options are permanent physical separation with no shared space at all, work with a qualified clinical animal behaviourist through your veterinarian, or rehoming one animal to a household without the other species. Management is a legitimate answer and the ASPCA frames it that way: in many cases the only solution is to manage the problem by limiting the dog's exposure to the situations that trigger her aggression, and pet parents should be prudent and always assume that their dog is not cured. A flat in Pune where the cat lives upstairs and the dog lives downstairs, permanently, is a worse outcome than friendship and a much better one than an emergency at midnight.

Resources and escape: height, an exit from every room, and bowls the other animal cannot reach

Most cat and dog households fail at the resources rather than at the introduction, because a cat with no exit from a room has only one move left and that move is the one that triggers chasing.

Four resources decide whether the pairing holds. The cat needs a safe place with vertical height it can reach and the dog cannot. It needs an escape route out of every room that does not pass the dog. It needs a litter tray the dog can neither block nor raid. And it needs to eat where the dog cannot reach the bowl. Fail any one of those and the introduction protocol is being run on a flat that cannot support it.

The Italian data shows how households actually solve the food half. Of 1,252 owners who answered the question, 42.2 percent said the dog ate from a bowl on the ground while the cat ate from a bowl positioned on top of something, 27.2 percent kept the two bowls far apart, 20.8 percent kept them close to one another and 3.9 percent fed both from the same bowl. The raised bowl is the commonest answer for a reason, and it is the right one in an Indian kitchen where there is no second room to feed in.

The failure mode has a number too. Asked what the animal that finishes its meal first does, 57.7 percent of owners said it moves away and 25.9 percent said it waits close to the animal still eating, but 7.3 percent, 88 of 1,207 households, said it removes the other animal and eats its meal. That is roughly one household in fourteen running a daily resource conflict at the bowl, and in almost every case the animal doing the removing is the larger one.

The bowl problem runs both ways, and both ways are nutrition problems

A dog eating cat food is not a funny habit, it is a dog eating a food built for a different animal. Cats are obligate carnivores, which means they rely on nutrients found only in animal products, and they evolved as hunters eating prey that contains high amounts of protein, moderate amounts of fat and a minimal amount of carbohydrates. Adult cats need 2 to 3 times more protein than dogs, cannot make their own arginine or taurine, need pre-made vitamin A because the enzyme that converts plant carotenoids is barely active in them, need the animal form of vitamin D, and need more niacin than other species. A cat food is therefore denser and richer than the dog's own ration, and a dog clearing it daily is being over-fed on top of its own meals.

The reverse matters more. These special needs are exactly why cats should not be fed dog food, and a cat that has lost its bowl to the dog and is eating the leftovers from the dog's bowl is, over months, on a diet missing taurine, which in deficiency causes retinal degeneration and an enlarged heart that cannot pump properly. How much a cat should actually be getting, in grams and in rupees, is in our cat diet chart for India, and the milk question that Indian households reach for instead is answered in can cats drink milk.

The environment half of this, meaning how much vertical space an Indian flat has and where each resource goes, is set out in full in how to make your house cat friendly. Tray count, tray type and house-soiling are in litter training a cat in India. Nets and mesh, which matter here because a cornered cat heads for a window, are in indoor cat safety in India.

The Indian flat: no spare room, a grille door, terrace cats and four people feeding

An Indian flat changes five things about a cat and dog introduction, and the first is that stage one assumes a spare room most households do not have.

Every foreign page writes the separation stage as though a spare bedroom exists. In a two-bedroom flat in Mumbai, Pune or Hyderabad with a working couple, a child and a grandparent, it does not. The workable substitutes are a bedroom with an attached bathroom, which gives the cat a tray behind a door the dog cannot open, or the kitchen plus the utility balcony, which gives height, a tray and a feeding surface in one zone. Neither is ideal and both beat running stage one in a corridor.

The grille door is the one Indian fixture that helps rather than hinders. Most Indian flats carry a steel safety door outside the main door, and many carry internal grille partitions to a balcony. A grille lets two animals see and smell each other with no possibility of contact, which is exactly what stage four of the protocol asks for, and it is already installed and already load-bearing in a way a lightweight baby gate is not. We have found no published study of grille doors used this way, so treat this as a sensible use of a fixture you already own rather than as an evidence-based recommendation.

1. Use the grille door or an internal grille as the barrier for the sight stage, and keep a cloth over the lower half if the dog is fixating.
2. Run the separation stage in the room with the attached bathroom, so the cat's litter tray never sits in the dog's territory.
3. Feed the cat on a kitchen counter or the top of the fridge from day one, before the dog ever meets it, because moving a cat's feeding place later is its own stressor.
4. Block the stairwell and the terrace until the indoor relationship is settled, since a dog that is reliable indoors can still hunt on a terrace.
5. Expect a humid Mumbai or Chennai August to push the cat higher and a Delhi January to push it towards a sunny floor, and give it a raised option in both.
6. Tell every person in the house the same three rules, because in a joint family the cat is being fed by four people and the dog is being let into the cat's room by whichever one of them is nearest the door.

Community cats on the landing are a different problem from the cat in your flat

A resident dog in an Indian building meets cats it does not live with, every day, in the lift lobby, on the stairs and on the terrace. The only published survey of Indian attitudes to those cats reached 763 residents across 13 cities, from Ahmedabad, Mumbai and Nagpur in Maharashtra in the west to Kolkata and Bhubaneshwar in the east, Delhi and Lucknow in the north, and Chennai in Tamil Nadu, Bengaluru, Kochi and Vijayawada in the south. It records an estimated 9.1 million stray cats in India, found that 36.4 percent of respondents had ever owned a pet including 22.7 percent who had owned cats, that 36.7 percent had ever fed a stray cat, that 62.1 percent agreed the welfare of stray cats is important, and that 72.2 percent had never heard of trap-neuter-return. The authors call it a pilot study, the first of its kind in India, and note that their sample was more educated than the profile of Indian cities and so may understate negative attitudes.

What that means on your landing is simple. The cats your dog meets outside the flat are not going anywhere, they are fed by somebody in your building, and your dog is going to pass them on a lead twice a day for years. A resident dog that is being taught not to chase the cat indoors and is allowed to lunge at a community cat on the stairs is being taught the opposite lesson twice a day. Work the lead the same way in both places. The legal position on community animals and the people who feed them is covered in how to adopt a cat in India and in adopting a dog in India.

The monsoon matters for one practical reason. From June to September a cat that normally uses a balcony or a terrace loses it for weeks, which collapses its territory back into the rooms the dog occupies at exactly the time an introduction might otherwise be going well. If you have a choice of start date, start in the dry months.

How long does it take for a dog to get used to a new cat?

A dog takes weeks to months to get used to a new cat, and no published study has measured the distribution, so any page quoting you a confident figure is guessing.

The Indian veterinary chain page that ranks for this query puts it at a few days to several weeks, depending on their temperament and the introduction process, and adds that some pets adjust within days while others may take weeks or even months. That is not wrong. It is also not measured, and the page carries no study, no number and no denominator anywhere.

Here is what has actually been measured, which is less than you would hope. A blinded randomised trial in multispecies homes ran six weeks, two of baseline and four of treatment, and found undesirable interaction scores significantly lower during treatment than at baseline, with the change detectable from week 3 onwards. The Dogs Trust cohort compared puppies whose cat-dog data were collected before 12 weeks of age with those collected between 12 and 22 weeks, and found only desirable behaviour in 10.7 percent of the younger group against 4.9 percent of the older one. Feuerstein and Terkel only observed pairs where both animals were over 6 months old, on the reasoning that younger animals have not yet established a stable pattern of behaviour.

Put those three together and a defensible operational answer falls out, which we label as our reading rather than as a measured figure.

• Plan 6 weeks of staged work before you expect the dog to be reliably uninterested, because that is the window in which a controlled trial saw measurable change.
• Judge the pairing at 12 weeks rather than at 2, because a dog that is still fixating after three months is not mid-process.
• Expect a puppy to take less time and an adult dog with a history of chasing cats outdoors to take the longest, or never to get there American Society for the Prevention….
• Expect the cat to take longer than the dog to relax, since cats were reported uncomfortable daily to weekly almost three times as often as dogs.
• Count weeks of correct practice, not weeks since the animal arrived. A fortnight of daily five-minute gate sessions is worth more than two months of both animals loose and tense.

Four things lengthen it: an adult dog rather than a puppy, a cat that has never met a dog, a first meeting that was rushed or went badly, and a flat with no height and no exits. Three of those four are the same factors the published models identified Feuerstein N and Terkel J.

How you know it has worked, and what chronic stress in the cat looks like

You know it has worked when the cat chooses to be in the room with the dog, eats and uses the tray on its normal schedule, and has stopped tracking the dog across the floor.

Toleration is not a bond, and the published affiliative markers are specific enough to check. In the Italian survey, 42.8 percent of dogs licked the cat and 34.3 percent of cats licked the dog, 68.5 percent of pairs slept near each other at least occasionally and 10.7 percent always, and 62.4 percent played together, of which 58.1 percent chased each other, 43.9 percent of cats played with the dog's tail and 63.8 percent made ambushes. Those are bonded-pair behaviours. A pair that shows none of them and no hostility either is in the indifferent category, which is a legitimate finish.

The qualitative half of the Dogs Trust study describes the same thing as a continuum running from friendship through an amicable relationship and separate lives to conflict and fear, and makes one point sharply: owners who let the animals introduce themselves generally expected the pair not to be friendly, and accepted behaviours such as aggression towards the dog, vocalisation, hiding and changes in general patterns of behaviour, which are typically indicative of signs of stress in cats. In other words, a lot of households file a chronically stressed cat under personality.

The reason the fifth row is a veterinary row rather than a behaviour row is that stress in cats is not only a mood. Urinary signs, straining in the tray, going outside the tray and blood in the urine sit on a disease that has environmental stress as a proposed mechanism, and we do not re-explain it here because it has its own page. Feline lower urinary tract disease covers the signs, the emergency and the treatment, and the one line worth carrying across is that a male cat straining in the tray and producing nothing needs to be seen the same evening.

What this page can prove, and what it cannot

Six claims on this page are measurements with denominators attached and six common beliefs about cats and dogs are not established, so the line between them is drawn here rather than left to the reader.

Six claims here are established, and each names the study behind it.

• Roughly two thirds of cat and dog pairs living together are amicable, with 66.5 percent of dogs and 65.5 percent of cats in 170 households, and under one in ten aggressive.
• The cat, not the dog, is the main controller of amicability, and cat factors outweighed dog factors in a regression on 748 homes.
• A gradual rather than immediate introduction raised the odds of only desirable puppy behaviour to 1.86, CI 1.14 to 3.05, in a model of 1,201 puppies.
• Cats injure dogs about ten times as often as dogs injure cats, at 9.6 percent of the cats against 0.9 percent of the dogs across 748 mixed-species homes.
• Cat claw injuries to a dog's eye perforate the cornea and lens capsule and ended in phthisis bulbi or secondary glaucoma in 50 percent of one hospital series, with puppies over-represented because they have no menace response before 12 weeks.
• An adult cat needs 2 to 3 times more protein than a dog and cannot make its own taurine, which is why the shared bowl is a nutrition problem and not only a conflict.

Six further claims are not established, and this is every place the page reasons rather than reports.

• No study of cat and dog introductions has been run in India. Every behavioural figure here comes from Israel, the United Kingdom or Italy, and the only Indian evidence on this page is a pilot survey about community cats.
• All three relationship surveys are owner reports, not observations, with voluntary samples and, in the Italian case, 91.2 percent women respondents.
• Nobody has measured how long a cat and dog introduction takes. The six-week and twelve-week figures on this page are our reading of a six-week trial and a puppy cohort, not a measured distribution Kinsman RH, Owczarczak-Garstecka SC.
• No validated test predicts which dog will kill a cat. The signals listed above are drawn from a welfare body's description of predatory behaviour, which carries no denominator.
• The pheromone trial that improved cat and dog interactions had no placebo arm, both arms improved, and its own authors asked for a four-group placebo-controlled replication. Bonbeno names no brand to buy and that trial is not a recommendation.
• The grille door as an introduction barrier is reasoning from a fixture most Indian flats already have, not a published finding, and we have said so where it appears.

One last boundary. Nothing here replaces an examination of your animals. If the dog has already caught the cat, if the cat has a puncture wound anywhere, or if either animal has an eye that looks different from the other one, that is a veterinary appointment today rather than a page to read. Where a dog is doing well with the cat and badly with everything else, the training side is in puppy training basics in India, and the flat-specific side of living with a dog is in keeping a dog in an Indian flat.