How much water should a dog drink, and a cat: the measured dose, the food that supplies most of it, and the Indian summer bowl

How much water should a dog drink, in millilitres

A healthy adult dog drinks and eats its way to roughly 60 to 80 ml of water per kg of bodyweight a day, and the honest version of that sentence carries a range so wide it changes what you do with it. The useful number is not the population average. It is your own dog's normal.

Start with the best measurement. Sixteen healthy adult dogs fed a dry kibble diet and given tap water were weighed in and out for eight weeks, with every bowl and bucket weighed by hand several times a day. Their total water intake, counting what they drank, the moisture in the food and the water their own metabolism produced, averaged 71 plus or minus 8 ml per kg a day, varying between 60 and 82 ml/kg from one day to the next. The authors note that this sits inside a previously reported mean of about 63 to 73 ml/kg/day for dogs drinking tap water on dry food.

Now the scatter, which no page ranking for this query prints. A review of canine polydipsia records that a wide range of water consumption, from 4 to 109 ml/kg/day, has been reported in normal laboratory dogs. That is a twenty-seven-fold spread inside the word normal. Against a spread like that, a page that tells you a dog needs 50 ml/kg is not giving you a target. It is giving you the middle of a distribution and hoping your dog is standing on it.

The reference range most people quote comes with a condition attached that is almost never repeated. The Merck Veterinary Manual's wording is that in a thermoneutral environment, most mammalian species need about 44 to 66 ml/kg bodyweight. Thermoneutral means an ambient temperature at which the animal spends no extra energy keeping warm or cooling down. A flat in Delhi in May, a terrace in Ahmedabad in June or a humid afternoon in Chennai is not thermoneutral, and the figure was never meant to describe one.

Why the per-calorie figure is the better rule, and where it came from

Express the requirement per calorie rather than per kilogram, because calories track what the animal is actually doing. The commonly cited ratio is 1.0 ml of water per kcal of metabolisable energy eaten, and the dogs in the eight-week study ran a weekly mean of 1.2 to 1.4 ml per kcal while holding a urine specific gravity between 1.025 and 1.030. The authors' own reading is that the conservative estimate underestimates their water requirement, and they say why: the references behind that estimate are generalisations from book chapters and not specific peer-reviewed research reports.

Read that twice, because it is the single most important sentence in this literature. The round number every website repeats was never a measurement. It was a cautious floor written into a reference text, and when somebody finally weighed the bowls, real dogs drank 20 to 40 percent more than it.

Measure your own dog's normal, which takes a week

To find your own dog's normal, fill a measuring jug each morning, pour from it into the bowl all day, and write down what is left at night. Do it for seven days, in the same season, and average it. That number, in millilitres divided by your dog's weight in kg, is the only figure that will ever tell you something useful, because what matters clinically is a change from an animal's own baseline rather than its distance from a population mean.

The Indian page that tells a 6 kg dog to drink 1500 ml

Check any water table against its own stated rule before you follow it, because the table ranking on page one in India fails that check by a factor of five. This is the clearest example of why a size band is the wrong unit for this question.

Purina India's hydration page gives a general guideline of 40 to 60 ml per kg of bodyweight a day, and then prints a table by size band: a toy dog of up to 2.5 kg, 125 ml; a small dog of 2.5 to 6 kg, 300 ml; a medium dog of 6 to 30 kg, 1500 ml; a large dog of 30 to 60 kg, 3000 ml.

Every one of those four figures is exactly 50 ml per kg calculated at the top of its band. 2.5 kg times 50 is 125. Six times 50 is 300. Thirty times 50 is 1500. Sixty times 50 is 3000. The table is internally consistent, and it is consistent in the one way that makes it useless, because almost no dog sits at the ceiling of its band.

Work it for the dogs Indian households actually keep. A 6 kg Pug or a small Indie falls in two bands at once, since the small band ends at 6 kg and the medium band starts at 6 kg. Read as small it gets 300 ml. Read as medium it gets 1500 ml, which is 250 ml per kg, between four and six times the page's own 40 to 60 rule. A 10 kg Indie read as medium gets 1500 ml, or 150 ml per kg, two and a half times the top of that rule.

This arithmetic is ours, performed on two figures that page publishes itself, and we are printing it rather than quietly disagreeing. The page is not dishonest. It has taken a reasonable per-kilogram rule and flattened it into bands for readability, and the flattening destroyed it. Any table that gives one answer for every dog between 6 and 30 kg has thrown away the only variable that matters.

Most of the water arrives in the food, and almost nobody counts it

Food, not the bowl, is where the largest share of a pet's daily water comes from on anything other than a dry diet, and the size of that share is the most useful number on this page. It has been measured in dogs, in grams, and the gap is thirty-four-fold.

Total daily water intake has three inputs: food moisture, free water, which is the term for everything the animal drinks, and metabolic water, the small amount produced when the body burns protein, fat and carbohydrate. Food moisture is the largest of the three on any diet that is not dry, and it is the one no owner counts.

Ten adult dogs were crossed over between a dry kibble diet analysed at 6.1 percent moisture and a fresh cooked food analysed at 71.1 percent moisture, with every gram of food and water weighed daily. On the kibble they ate 180.5 g of food a day and took 11.0 g of water from it. On the fresh food they ate 528.0 g a day and took 375.4 g of water from it.

The dogs did try to compensate at the bowl, and they did not manage it. Free drinking ran 423 g a day on the kibble against 147 g on the fresh food, so the dogs on dry food drank nearly three times as much. Even so, total daily water intake finished higher on the fresh food, at 522.5 g against 434.1 g (p < 0.0001), and the fresh-fed dogs met 141 percent of their calculated minimum water requirement against 102 percent for the dry-fed dogs.

The 102 percent figure is the one to sit with. A dog on dry food, drinking freely from a full bowl in a temperature-controlled room, cleared its minimum requirement by two percent. There is no margin in that at all, and the Indian summer is precisely the condition that eats margins.

The Indian home-cooked bowl is a wet diet

An Indian dog eating rice at home is on a high-moisture diet whether or not anybody has called it one, and this is the part no page written for India has worked out. Cooked white rice holds 68.4 g of water per 100 g. That sits just below the 71.1 percent fresh food and the 75.1 percent canned food in the trial above, and far above anything that comes out of a bag.

The ceiling on the bag is set in Indian law. Dry dog food supplied against the Bureau of Indian Standards specification IS 11968:2019 must hold a maximum of 10 percent moisture by mass, a figure written into the purchase specification the Thiruvananthapuram City Mounted Police Unit issued for its own working dogs, signed by its veterinary surgeon. The Merck manual's observed range for dry pet foods generally is 3 to 11 percent, with canned foods running 60 percent to more than 87 percent.

So the practical reading for an Indian household is straightforward. A dog eating a rice-based home-cooked bowl in Chennai, Kolkata or Bengaluru is behaving like the fresh-food arm of that trial and will drink visibly less from its bowl, and that is normal rather than worrying. A dog switched from that bowl to kibble, which is what happens in a great many Indian homes during a move or on a vet's advice, has just had roughly nine tenths of its dietary water removed and has to make the whole difference up by drinking. Nobody tells the owner that, and the owner notices only that the dog is suddenly at the water bowl all day. What a home-cooked Indian bowl does and does not deliver on everything other than water is on our guide to homemade dog food in India, and the portion arithmetic for rice is on can dogs eat rice.

Where the studies disagree, printed rather than averaged

The fresh-food result contradicts a good deal of earlier work, and the authors say so themselves. Their introduction records that these results contradict previous studies that demonstrated total water intake did not differ in dogs fed a dry or canned diet, or that dogs fed canned diets actually consume less total water due to a reduction in free-water intake. The classical position is that a dog adjusts its drinking to hold total intake roughly constant whatever the food's moisture, and several studies support it.

Both positions can be true at once, and the mechanism is palatability. A dog that eats 528 g of something it likes has swallowed 375 g of water before it thinks about drinking, and it would have to actively refuse water at the bowl to end up level. A dog fed a diet it eats less enthusiastically does not. In the same paper's second study, dogs offered a canned diet finished only 79 percent of what they were given, against 93 percent of the fresh food. Moisture only reaches the animal if the animal eats the food.

Cats barely drink, and that is normal for the species

A cat drinks far less than a dog of the same weight because it evolved to take its water in its prey, and the practical consequence is that a cat's diet decides its hydration far more than its bowl does.

The commonly stated requirement is about 50 ml of water per kg of bodyweight a day, which is 200 to 250 ml for a cat of 4 to 5 kg. A cat's natural food, small mammals and birds, runs about 70 percent moisture, so at the 200 to 250 g of prey that covers its energy needs, roughly 70 percent of a cat's fluid requirement is already covered by eating, before metabolic water is counted at all. On an 80 percent moisture wet food the requirement can be covered completely. On commercial dry food, which holds a maximum of about 10 percent moisture, the cat must drink the difference.

Two pieces of feline physiology sit underneath that. Cats tolerate acute fluid losses of up to 20 percent of bodyweight comparatively well, and they can produce highly concentrated urine to conserve body fluid. A cat is built to run dry without complaining, which is exactly why a cat that is running dry does not look like it is.

The desert-cat story, and the caveat its own source prints

You will read everywhere that cats drink little because they descend from a desert animal. The veterinary account of this is more careful than the internet version, and worth repeating honestly. It has been suggested that these abilities come from the domestic cat being a descendant of the African wildcat, an alleged desert dweller; but the same authors immediately note that domestication most probably began 9,000 to 10,000 years ago in the fertile crescent, which was by no means a desert at that time.

So treat the desert ancestry as a plausible explanation that the literature itself hedges, and treat the measured behaviour as the fact. The cat drinks little. Why it drinks little is less settled than the pages telling you about it suggest.

Low intake is linked to urinary and kidney disease in cats, and that argument is made in full, with the Indian clinic context, on our pages for feline lower urinary tract disease and chronic kidney disease in cats. This page does not rebuild it. What belongs here is the one line that follows from the arithmetic above: the diet, not the bowl, is the lever with the largest measured effect, and the full feline case for wet food in an Indian summer is on heat safety for cats in India.

How to tell if an animal is actually dehydrated, and what the skin tent is worth

To tell whether an animal is dehydrated, weigh it and count what it drinks, because the three signs owners are taught to use have been tested against a gold standard and two of them failed. This is the section of the page most likely to change what you do.

The signs of dehydration every pet page lists are the skin tent, dry or tacky gums and a slow capillary refill time. All three are physical examination findings borrowed from the clinic, and the question nobody asks is how well any of them performs when the animal is not already obviously ill.

The skin tent is the famous one. You lift a fold of skin and time how long it takes to fall back. It is on every pet page in existence and it has been properly tested once, in a study designed to validate it.

Fifty working horses with a positive skin tent test, working in environmental temperatures of 30 to 44 degrees C in Pakistan, were rested and offered water freely. Bodyweight, blood parameters, mucous membrane dryness, drinking behaviour and skin tent duration at six anatomical locations were measured at seven time points across five hours, with plasma osmolality as the gold standard for dehydration.

Skin tent duration was affected by which side of the animal you tested (p = 0.008), by the anatomical location and by how wet the coat was (both p < 0.001), and younger animals had shorter skin tents at every time point (p = 0.007). There was no significant association between plasma osmolality or water intake and skin tent duration. The authors' conclusion is unambiguous: the standardised skin tent test and measures of mucous membrane dryness investigated in this study were not valid or repeatable indicators of dehydration when compared with plasma osmolality as a gold standard criterion.

They also say what did work. The volume of water consumed and the number and duration of drinking bouts were the most reliable guide to hydration status currently available, and offering palatable water freely provides both the diagnosis and the remedy. An animal that drinks a great deal when water is offered was short of water. That is the test.

Where the dog study and the horse study disagree

The one controlled test in dogs reached the opposite conclusion, and both are printed here rather than averaged. Nine exercise-conditioned working dogs were measured before and after a 15-minute exercise period. They lost a mean 0.83 plus or minus 0.27 percent of bodyweight, a range of 0 to 3.3 percent, which was not a statistically significant loss. Skin tent time nevertheless increased significantly after exercise, both as a real-time measurement in the field (p = 0.006) and on later video review (p = 0.05), and the two readings correlated (r = 0.68, p < 0.001). The authors concluded that monitoring skin tent time appeared to be a useful strategy for predicting small shifts in hydration status of dogs during exercise.

Read the two together and the resolution is about what each was asking. The dog study asked whether skin tent time moves when a known, mild fluid loss occurs in the same nine animals, and it does. The horse study asked whether a single skin tent reading tells you how dehydrated an individual animal in front of you actually is, and it does not, because the reading depends on which side you pinch, where you pinch, how wet the coat is and how old the animal is. Those are different questions and the second one is the one an owner is asking at home.

Capillary refill time comes off worse than the skin tent in both. In the dog study it decreased after exercise rather than increasing, the change was not significant, and the authors state plainly that it was not a significant predictor of a change in hydration status.

Getting water in: what is measured and what is folklore

To get more water into a pet, change the moisture of the food first and the flavour of the water second, because those are the two interventions with measured effects and the bowl-shaped ones mostly do not have any.

Flavouring the water is the intervention with the clearest number behind it, and it was measured in conditions an Indian owner will recognise. Seven detection dogs worked across study days with a median environmental temperature of 29.3 C (range 23.3 to 37.7 C) and a median humidity of 70 percent. Offered plain water, they drank a mean 7.04 plus or minus 3.42 ml/kg per hour. Offered a flavoured oral electrolyte solution, they drank 15.61 plus or minus 4.47 ml/kg per hour. A third group given subcutaneous fluids drank 5.56 plus or minus 4.40.

That is a doubling of voluntary intake from making the water taste of something, in heat and humidity that match Mumbai or Chennai in April. It is the single most actionable finding on this page for anyone with a pet that will not drink.

1. Offer more stations than you think are needed. Multiple water sources encourage consumption, and the Merck manual flags this as particularly important in cats, which often do not drink a lot of water.
2. Flavour one of them. The measured effect was a doubling of drinking in working dogs in heat. Plain cooking water off poached chicken, with no salt, no masala and no onion, is the household version.
3. Add water to the food rather than only to the bowl, since dietary moisture is the lever with the largest measured effect.
4. Keep bowls small and shallow for a cat. In a survey of 549 cat households, bowls under 15 cm in diameter were preferred to larger ones, and no preference was reported for the material the bowl was made of.
5. Move the water away from the food. In that survey 41 percent of households put the water bowl right next to the food bowl, while cats given a choice preferred water points in a different room entirely.
6. Refill rather than top up, for the reason set out in the next section.

Fountains, and the one cat in twelve that hated it

Water fountains are recommended on almost every cat page and the controlled evidence for them is weak. In the 549-household survey, more than 80 percent of cats were given water in bowls with the fountain as the commonest alternative, and among cats that had access to both options the majority preferred the bowl. The authors note this is consistent with other studies that did not find a statistically significant difference in water intake between bowls and fountains, finding instead large individual differences and preferences.

One detail from that literature deserves to be better known than it is. In a study of twelve cats, one cat was stressed by a water fountain to the extent that it showed aggression, overgrooming and vomiting. A fountain is a reasonable experiment on one animal and it is not a guaranteed improvement, and it can make things worse for roughly one cat in twelve. The randomised crossover trial that tested still, circulating and free-falling water in cats is set out in full on our can cats drink milk page, and we are not repeating it here.

Ice is the other popular suggestion, and it is a preference question rather than a hydration one. Offer a bowl with ice and a bowl without and let the animal pick. Nothing is lost if it ignores the ice.

The bowl itself, and what grows in standing water

A standing water bowl builds a bacterial film within days whatever it is made of, and the two studies that have measured it point in opposite directions on which material is worst. Scrub it daily and stop trying to buy your way out of the problem.

The Hartpury University study is the one that made the newspapers. Six clinically healthy medium-sized dogs were each given new, sterilised plastic, ceramic and stainless steel bowls, which were swabbed at day 0, day 7 and day 14 and cultured. Bacterial counts rose significantly for all three materials between day 0 and day 7 and between day 0 and day 14 (p < 0.05), with no significant difference between materials (p > 0.05), though plastic held the highest count descriptively at day 14. Several medically important bacteria were identified, including MRSA and Salmonella, with the majority of species coming from the ceramic bowl, which the author reads as suggesting harmful bacteria may develop biofilms more successfully on ceramic.

The larger study found something that contradicts the popular advice outright. Ninety-six dog food bowls from Italian households were sampled and analysed. Metal bowls carried 2.18 log10 CFU per sq cm of total mesophilic aerobic bacteria against 1.23 for plastic, and the difference was significant at p < 0.001. Bowls used for wet food carried 1.99 against 1.42 for dry food (p = 0.0397). No Salmonella, Campylobacter or verotoxigenic E. coli was detected in any of the 96 bowls.

The two studies point in opposite directions, and both are printed

One study says plastic held the most bacteria and ceramic the nastiest ones. The other says metal carried nearly ten times the load of plastic. They are not measuring the same thing: the first used new sterile bowls holding drinking water for 14 days in six dogs, the second sampled food bowls already in use in 96 real households with whatever cleaning habits those households had. Neither is wrong and neither supports the line, repeated on every retailer's page, that switching from plastic to stainless steel solves the problem.

What both support is that the bowl is dirty and that cleaning is the variable. The Italian team cite a hygiene protocol using hot water at 71.1 degrees C and soap after each use, which produced lower bacterial counts than washing with cold or lukewarm water. Their own data also found Enterobacteriaceae counts higher in hand-washed bowls than dishwasher-washed ones (p = 0.0515). Hot water and friction are doing the work, not the material.

• Empty and scrub the water bowl daily with hot water and detergent, not a rinse. A rinse moves biofilm around; it does not lift it.
• Replace the water rather than topping it up. Adding fresh water to a slimy bowl refills the bowl and reseeds it.
• Run the bowl through the hottest wash you have once a week, a dishwasher cycle where you have one.
• Feel the inside with your fingers before you refill. A film you can feel is a film the sponge has been skipping.
• Retire a bowl that is visibly scratched, in any material, because the scratches are where the film sits out of reach of a sponge.
• Treat added flavourings with the same care. The 2025 trial's authors warn that water supplements may, depending on the ingredients, predispose the water to additional contamination or bacterial growth.

What an Indian summer changes, and what it does not

An Indian summer raises water loss and therefore water intake, and it changes almost nothing about the arithmetic above except the scale. The thermal half of the problem, which is the dangerous half, belongs on another page.

The reference figure of 44 to 66 ml/kg is explicitly conditioned on a thermoneutral environment, and India spends much of the year outside one. The detection dogs that drank 7.04 to 15.61 ml/kg per hour were working at a median 29.3 C and 70 percent humidity. Those are ordinary conditions in Mumbai, Chennai or Kolkata rather than extreme ones, and the hourly rate they imply across a working morning dwarfs any daily table.

Humidity matters more than the thermometer in the coastal belt. A dog loses most of its heat, and a great deal of its water, through panting, and wet air blunts the evaporation that panting depends on. In Chennai, coastal Karnataka, coastal Maharashtra and the Kolkata delta the thermometer reads lower than inland and the animal copes worse. In dry inland cities like Delhi, Ahmedabad, Hyderabad and Pune the air takes water readily, which helps the dog cool and empties the bowl faster.

The Animal Welfare Board of India treats water provision as a duty rather than a kindness. Its summer advisory to every Chief Secretary and to compassionate citizens, animal owners and care takers states that animals are vulnerable to dehydration, heatstroke and burns from hot surfaces, and asks for shade through the day and water bowls kept full, framed under Section 3 of the Prevention of Cruelty to Animals Act, 1960.

The bowl, the tank and the power cut

Three Indian household facts change the practical job, and none of them needs a number to act on. Water drawn from an overhead tank in the afternoon comes out warm, so draw and store the day's drinking water in the early morning when the tank is at its coolest. A bowl left standing on a sunny floor warms toward room temperature within an hour, which both discourages drinking and speeds whatever is growing in it, so site bowls in the coolest, shadiest corner of the flat. And a power cut takes the fan and the cooler with it, which raises the animal's losses at exactly the moment nobody is topping the bowl up, so leave more water out than you think is needed before you leave the house in summer.

The AWBI's own answer for a balcony, a gate or a street is a wide earthen pot in permanent shade, refilled morning and evening. An earthen matka holds water cooler than steel through evaporative loss from its own wall, which is the same physics that cools a matka of drinking water in an Indian kitchen, and it is heavy enough that a dog cannot tip it.

Two other seasons move the number in the opposite direction and are worth a line each. Through the monsoon, humidity stays high while temperatures fall, intake drops and the bowl grows a film faster in damp air; the rest of that season's problems are on monsoon care for dogs. Through a north Indian winter, dogs drink less in cold, dry air at exactly the point they should not, which is covered in winter care for dogs in India.

Drinking much more is a symptom, not a success

A sudden rise in thirst is a clinical sign that sends a veterinary surgeon looking for disease, so treat an increase in your dog's own normal as a reason to book an appointment rather than as proof the hydration advice worked.

Polydipsia in dogs is defined as fluid intake that exceeds 90 to 100 ml/kg/day. Translate it: a 10 kg Indie crossing a litre a day, every day, is at the line. A 25 kg Labrador crossing two and a half litres is at the line. Those are not hard numbers to reach in May, which is exactly why the measurement has to be made against the animal's own baseline and across several days rather than one hot afternoon.

The threshold itself is disputed, and the disagreement is useful. One study found the upper limit of normal daily water consumption to be approximately 70 ml/kg/day, quite a bit less than the commonly used 100 ml/kg/day figure, and the review that records this warns that using a specific water intake cut-off could be misleading in those dogs who do not reach 100 ml/kg/day. Its own advice to clinicians is that owners can be instructed to measure water intake at home over a period of several days, and that polydipsia may manifest itself at much lower daily volumes in dogs who normally drink little.

That is the whole argument for the measuring-jug week in the first section. A dog whose baseline is 40 ml/kg and which now drinks 85 has doubled its intake and will never trip a 100 ml/kg threshold. The owner who wrote the baseline down in February is the only person who will catch it.

• Book an appointment for a sustained rise in drinking, alongside more urine, in an animal whose food and weather have not changed.
• Mention the number. Say what the animal drank, in millilitres, over how many days, because that is the figure the clinician would otherwise have to ask you to go home and collect.
• Expect a urine test early. Urine specific gravity is the cheap first measure of whether the kidneys are concentrating, and it is routinely used alongside water intake.
• Read the condition pages rather than this one for the diseases behind it. Increased thirst is an early sign in diabetes in dogs and cats and in chronic kidney disease in cats, and low intake sits underneath bladder stones in dogs and feline lower urinary tract disease.

Water intoxication, the opposite problem

Too much water swallowed too fast is a genuine emergency and it is rare. Cornell's canine health centre describes the mechanism plainly: a dog in and out of a pool or a lake all day, especially chasing balls or sticks, can ingest a lot of water while playing, and when dogs ingest too much their electrolytes in the bloodstream become diluted, which can cause brain swelling that, if left unchecked, leads to seizures and other neurological derangements, possibly death. Their prevention is to make sure the dog takes frequent half-hour breaks from the water through the day.

Sea water does the reverse and matters on an Indian beach. When a dog consumes too much saltwater, excess salt in the bloodstream pulls water out of brain cells, with muscle tremors, disorientation and seizures as the signs; Cornell's prevention is to send the dog to the beach already well hydrated so it is less likely to drink sea water out of thirst, and to keep fresh water available there. What else a Goa or Chennai beach day asks of you is on taking your dog to the beach in India.

Electrolytes and ORS: what has actually been tested in dogs

An oral electrolyte solution formulated for dogs has been tested in dogs and worked; a sachet of human ORS from an Indian chemist is a different product designed for a different animal, and a dehydrated pet is a veterinary case either way.

Take the veterinary evidence first, because it exists. Twenty dogs with haemorrhagic diarrhoea and fewer than three episodes of vomiting were offered an oral electrolyte solution rather than going straight onto a drip. Thirteen of the twenty, 65 percent, voluntarily consumed it, all within five hours of admission; the other seven refused it and were given intravenous fluids instead. At 8 and 16 hours the oral group's packed cell volume, total protein and blood urea nitrogen were significantly lower than at admission, while venous blood pH, base excess and the concentrations of sodium, potassium, chloride, ionised calcium, ionised magnesium and lactate showed no significant change.

The authors' conclusion is that rehydration therapy with an oral electrolyte solution was effective and safe in dogs with mild to moderate dehydration associated with haemorrhagic diarrhoea, and that it cost significantly less than the intravenous alternative. Note three things about that trial. The animals were in a hospital. They were screened for parvovirus and given an anti-emetic first. And a third of them simply refused to drink the solution, which is the practical ceiling on the whole approach.

Human ORS is a different formulation, for a different problem

The sachet in an Indian medicine cabinet is a defined product. The World Health Organization's reduced osmolarity formulation contains 75 mEq per litre of sodium and 75 mmol per litre of glucose, for a total osmolarity of 245 mOsm per litre, settled by an expert group after trials in adults and children with cholera and with acute non-cholera diarrhoea. It is an excellent product for the job it was designed for, which is a human being losing sodium-rich fluid through the gut.

A dog is not that patient, and sodium is where the caution sits. Dietary sodium has a direct, linear correlation with how much a dog drinks, and raising it may be contraindicated depending on health status, specifically in cardiac disease and in calcium oxalate urolithiasis. The flavoured solution the detection dogs drank delivered an average of 416 mmol of additional sodium over a working day, which is a deliberate dose given to working animals under supervision, not a routine top-up. A sugar load is its own question in a diabetic animal.

What is safe at home is the unglamorous half. Offer palatable water freely, which is both the diagnosis and the remedy for a healthy animal that is merely short of a drink. Healthy animals given ample water self-regulate their intake effectively, and intake should not generally be restricted. A dog that is well and thirsty will fix itself from a clean bowl. A dog that is ill will not, and that is the line.

What this page can prove, and what it cannot

Seven claims on this page rest on measurements with denominators and four common claims about pet hydration are not established, so the line between them is drawn here rather than left for you to find.

• Measured total water intake in 16 adult dogs on dry food averaged 71 plus or minus 8 ml/kg/day, with a daily range of 60 to 82.
• Normal laboratory dogs have been reported across a range of 4 to 109 ml/kg/day, so the spread inside normal is enormous.
• Dogs on a 71.1 percent moisture diet took 375.4 g of water a day from their food against 11.0 g on a 6.1 percent kibble, and finished with higher total intake despite drinking far less.
• The standardised skin tent test and mucous membrane dryness were not valid or repeatable indicators of dehydration in 50 horses tested against plasma osmolality.
• Flavouring the water more than doubled voluntary drinking in working dogs at a median 29.3 C and 70 percent humidity, from 7.04 to 15.61 ml/kg/h.
• Bacterial counts rose significantly in new plastic, ceramic and stainless steel drinking bowls across 14 days, with no significant difference between materials.
• Dry dog food supplied to the Indian standard IS 11968:2019 carries a maximum of 10 percent moisture by mass.

Four things this page does not establish. Nobody has measured water intake in dogs or cats eating an Indian home-cooked bowl, so the inference that a rice-based bowl behaves like a wet diet rests on one moisture figure and one trial on a diet of similar moisture. Nobody has measured how fast biofilm builds in a bowl in Indian summer heat, and both bowl studies ran in temperate Europe. The desert ancestry of the cat is offered in the veterinary literature as a suggestion that its own authors hedge, not as a demonstrated cause of low thirst. And the threshold at which drinking becomes polydipsia is itself disputed, with one study putting the upper limit of normal at about 70 ml/kg/day against the 100 ml/kg/day in common use.

One further boundary. A measured daily water intake is a number, not a diagnosis. Everything this page says about increased thirst, dehydration and electrolytes exists to get you to a clinic sooner with better information, which is why it publishes under the medical disclaimer at the foot of the page.