Eight glasses a day is not a finding
The most repeated hydration advice in the English language has no source.
In 2002 the physiologist Heinz Valtin went looking for the evidence behind "drink at least eight 8-ounce glasses of water a day" and published what he found in the American Journal of Physiology: nothing. No study, no expert panel, no dataset. What he did find was a plausible ancestor — a 1945 US Food and Nutrition Board recommendation of about 2.5 litres a day, which continued, in the same paragraph, that most of that quantity is already contained in prepared foods.
That second sentence is the one that disappeared. Strip it out and a total-water figure that includes soup, fruit, coffee and yoghurt becomes a plain-water target roughly a third higher than anyone intended.
Valtin's conclusion was not that hydration is unimportant. It was that healthy adults in a temperate climate, eating normally, do not need to drink to a schedule to stay hydrated. Thirst works.
What the reference bodies actually say
Two sets of official figures matter, and they disagree by a wide margin — which is itself informative.
EFSA (2010) sets an adequate intake of 2.0 litres a day for women and 2.5 litres for men, adults, under conditions of moderate ambient temperature and moderate physical activity.
The US National Academies (2005) set 2.7 litres for women and 3.7 litres for men, and note that about 80% of intake typically comes from beverages and 20% from food.
Both are total water — everything you drink plus everything in what you eat. Neither is a plain-water target. And both are adequate intakes rather than requirements: an AI is the observed median intake of a healthy population, adopted precisely because the evidence was not strong enough to set a requirement. They describe what healthy people drink, not what they must drink.
The gap between the European and American numbers is roughly 35%. Two expert panels, similar populations, the same literature, and no agreement on the answer to a factor of a third. That is the honest state of the evidence, and it is why this calculator shows both alongside its own estimate rather than pretending to a precision nobody has.
What this calculator does
It works from body weight, which scales with the surface area and metabolic rate that actually drive fluid turnover, rather than from sex alone.
Baseline total water = weight in kg × 33 ml
Thirty to thirty-five millilitres per kilogram is the range in ordinary clinical use for adults; the tool uses the midpoint. A 70 kg adult gets about 2.3 litres, which lands between the EFSA and National Academies figures — as it should.
Three adjustments follow:
- Exercise. Minutes of activity multiplied by a sweat rate: 0.4 L/h for light activity, 0.8 for moderate, 1.2 for vigorous. Added on top, not scaled.
- Climate. Warm conditions add 10% to the baseline, hot or humid conditions 20%. This applies to the baseline only — the exercise term already carries its own sweat rate, and applying heat twice would double-count it.
- Pregnancy and breastfeeding. EFSA adds 300 ml a day during pregnancy and 700 ml during established breastfeeding.
Finally, about a fifth of the total is subtracted as food water to give a drinks target, following the National Academies' split.
Worked example
A 68 kg woman, 45 minutes of moderate running, warm weather:
- Baseline: 68 × 33 = 2.24 L
- Climate: 2.24 × 1.1 = 2.47 L
- Exercise: 0.75 h × 0.8 = 0.60 L
- Total: 3.07 L, of which about 2.5 L from drinks — roughly ten 250 ml glasses.
Compare that with her EFSA adequate intake of 2.0 L. The difference is almost entirely the run and the weather, which is the point: the official figures assume moderate conditions, and most people's days are not moderate.
Sweat rates vary more than anything else here
Every number above is a population average sitting on top of the largest source of individual variation in the whole calculation.
Measured sweat rates in adults range from about 0.5 to 2.0 litres per hour, and occasionally beyond 3 in heat-acclimatised athletes working hard in humidity. Two people of the same size doing the same session in the same room can differ fourfold. No formula can know which one you are.
There is, however, a direct measurement, and it takes one session:
- Weigh yourself with minimal clothing immediately before exercise.
- Train, noting how much fluid you drink.
- Weigh yourself again immediately after, towelled dry.
Weight lost plus fluid drunk equals fluid lost. One kilogram of body weight is very close to one litre. Divide by the hours trained and you have your own sweat rate — for that intensity, in those conditions. Repeat it once in the cold and once in the heat and you know your range, which is more than any calculator can tell you.
The American College of Sports Medicine's position is to limit losses during a session to under about 2% of body weight, and to replace the deficit afterwards rather than trying to match losses drink-for-drink during hard exercise, which is often impossible and sometimes counterproductive.
Everything you drink counts
Tea and coffee count. The diuretic effect of caffeine at ordinary doses is real but small, and nowhere near large enough to offset the water in the cup — a mug of tea is a net gain of most of a mug of water. The NHS counts tea and coffee towards daily fluid intake, and so should you.
Milk, juice, squash and soft drinks all count as fluid too, though the sugar in several of them is a reason to moderate them for other reasons entirely.
Alcohol is the real exception. Ethanol suppresses vasopressin, the hormone that tells the kidneys to conserve water, so beer and wine produce a net fluid loss at typical drinking volumes. That, plus its effect on sleep, is most of what a hangover is.
Food contributes more than people expect. Cucumber, watermelon, lettuce and tomatoes are over 90% water. Soup, yoghurt, pasta and rice are all substantially water by weight. A person eating a normal mixed diet takes in 700 ml to a litre a day without drinking anything.
Too much is a real risk, not a rhetorical one
The kidneys can excrete roughly 0.7 to 1.0 litres per hour. Drink faster than that for long enough and blood sodium falls, cells swell, and the result is hyponatraemia — headache, nausea, confusion, and in severe cases seizures, coma and death.
It is not hypothetical. Exercise-associated hyponatraemia is well documented at marathons and in military training, and it is caused specifically by over-drinking plain water while losing sodium in sweat. Slower runners drinking at every station are the classic case, because they have both the time and the opportunity.
The practical rules that follow: do not force fluid beyond thirst during long endurance events, use a drink containing sodium for sessions over about two hours, and treat "drink as much as you can" as bad advice rather than cautious advice.
The two checks worth more than a target
Urine colour. Pale straw means you are broadly fine. Dark yellow or amber means you are behind. It is crude, free, and available all day. Vitamin B supplements turn urine bright yellow and break the signal for a few hours; nothing else common does.
Thirst. For healthy adults under about 65, thirst is a well-calibrated signal — it responds to plasma osmolality, which is precisely the variable that matters. It lags during hard exercise, and it blunts with age, which is why drinking to a schedule is commonly advised for older adults, especially in hot weather or during illness.
If both of those look right, the exact litre figure does not matter much.
Where this estimate stops
It assumes a healthy adult with normal kidney, heart and liver function. Several conditions come with a fluid restriction — advanced heart failure, some kidney disease, cirrhosis with ascites, SIADH — and in those cases a higher intake is not merely unnecessary, it is harmful. Diuretics, lithium and some other medicines change the arithmetic too. If a clinician has given you a limit, that limit is the answer and this page is not.
It does not model altitude, which raises fluid loss through faster breathing and drier air; illness with fever, vomiting or diarrhoea, which can raise losses by litres a day; or occupational heat exposure, where industrial hydration guidance is stricter than anything here.
It also does not know your sweat rate, and that is the single biggest lever in the whole calculation. Weigh yourself before and after one session, and you will learn more about your own hydration than this or any other calculator can tell you.
Nothing here is medical advice. The formulas are population estimates, and a target that feels wrong for you probably is.