← About

How much water do you need?

Updated

This page sums up what research actually says about drinking water. Every claim links to a source in the list at the bottom; tap a [number] to jump to it.

The short version: there is no single number that fits everyone. The official values are reference points for healthy people in a temperate climate, and for most healthy adults thirst is a good guide.

What official bodies recommend

The European Food Safety Authority (EFSA) sets the adequate intake at about 2.0 litres a day for women and 2.5 litres for men [1]. The US Institute of Medicine (IOM) arrived at higher values: 2.7 and 3.7 litres [2]. Both figures describe total water — everything you drink, plus the water contained in food.

Food contributes roughly 20–30% of that total [1], so the part you drink is about 1.4–2 litres going by EFSA. The IOM measured the split directly in a national survey and put food at about a fifth, which leaves roughly 2.2–3 litres [2]. The two bodies studied different populations with different methods — treat the spread as an honest range rather than a contradiction. Little Often’s default goal — 2 litres (68 fl oz) of drinks a day — is where the EFSA and IOM ranges meet: the top of EFSA’s share from drinks, just under the IOM value for women; adjust it to your body, climate and activity.

“8 glasses a day” and “X ml per kg”

The “8×8” rule probably traces back to a 1945 recommendation whose second half — that most of this water is already contained in food — was simply forgotten; physiology and medical reviews class it as a myth [3] [4] [5]. Formulas like “30–35 ml per kg of body weight” come from clinical fluid therapy and from observed-intake reference tables (such as the D-A-CH values used in Germany, Austria and Switzerland [6]); they were never validated as a drinking requirement for healthy adults.

As rough starting points they are harmless — just know they are rules of thumb, not scientific norms.

Thirst is a good guide

For healthy adults under everyday conditions, the official reports say it plainly: you can rely on thirst [2] [7]. Needs do rise with heat and exercise, in pregnancy (about +300 ml a day) and while breastfeeding (about +700 ml) [1].

How much water people actually use varies enormously: in the largest study to date — 5,604 people in 23 countries — adult water turnover ranged from about 1.5 to 6 litres a day, and the authors concluded that one figure for everyone does not exist [8]. Turnover is not the amount you need to drink, though: food and the body’s own metabolism cover a good part of it, so drinking water is only part of the total [8]. A 2024 review found only 18 randomised trials of drinking more water, most of them small, and the benefits that did turn up stayed tied to particular conditions — weight loss, kidney stones, recurrent urinary tract infections in women who drank little to begin with [9]. So treat your daily goal as a personal reference point, not a medical target.

What colour should urine be?

There is also a rougher check that needs no counting: the colour of your urine. Pale yellow is what you are aiming for — the official figures themselves are built around moderately concentrated urine, not maximally dilute urine [1]. Consistently dark suggests you could drink more [10]; consistently colourless usually means you are drinking more than you need [1] [11] — colour tracks intake closely enough that drinking about a litre more per day visibly lightens it [11]. Read it as a trend rather than a measurement: some vitamins, foods and medicines change the colour on their own, and it says little in the hour after a large drink or hard exercise.

Tea and coffee count — and so does almost everything else

The most comprehensive head-to-head trial of common drinks tested 72 men, each drinking a litre of water plus up to three other drinks — 13 drinks in all — with urine collected for four hours [12]. None of them retained water worse than plain water — not coffee, tea, cola, diet cola, juice, sports drink or 4% lager. The differences between them were too small to detect — each drink was tested in about 17 men, enough to pick up a difference of roughly 200 ml per litre. Two retained more: milk and oral rehydration solution kept roughly 300 ml more of the litre in the body [12]. Independent studies of people rehydrating after exercise point the same way: milk and milk-protein drinks are retained better than water or a plain carbohydrate–electrolyte drink [13] [14], though the ~300 ml figure itself comes from a single industry-funded trial.

A later trial by the authors of that beverage comparison found that what changes retention is sugar and salt, not caffeine [15]. The salt half of that was established long ago and independently: table salt is sodium chloride, and it is sodium that keeps water in the body. Drink a lot of plain water quickly and blood sodium and osmolality fall, the body switches off its water-conserving hormones, and the surplus leaves as urine; replacing sodium along with the water prevents that [16] — the same principle behind oral rehydration solutions in medicine [17]. Most of your sodium comes from food, not drinks, which is why this matters mainly after heavy sweating rather than for the water you drink at the kitchen tap. The carbohydrate effect is concentration-dependent and partly reflects delayed stomach emptying rather than true extra retention [18]. Sugary drinks still hydrate, but they bring the sugar along. Moderate coffee drinking does not dehydrate you [19] [20] — and the same conclusion comes from a meta-analysis with no industry funding declared: across 16 studies, about 300 mg of caffeine produced roughly 100 ml of extra urine, mostly in women, and the effect vanished with exercise [21].

This is why Little Often has no drink-type picker: for everyday non-alcoholic drinks the volume is the only number that matters, so a cup of tea goes in as a cup. Milk is logged as its volume too — the extra retention was measured over four hours after an unusual litre of milk, and nobody has shown that a glass is “worth” more over a whole day.

Does alcohol count?

Alcohol is the one thing not to log. Weak beer (around 4%) came out close to water in these trials [12] [22], though urine output does creep up with alcohol strength [23]; wine and spirits make you pass clearly more urine for a few hours after drinking [22], and the evidence beyond that window is thin.

Can you drink too much? Yes

Healthy kidneys can excrete at most about 0.7–1 litre per hour [2] [24]. Drinking noticeably faster than that for hours dilutes the sodium in the blood — water intoxication (hyponatremia). It is rare, but dangerous, and it is why “the more the better” is bad advice [7].

The pace matters more than the daily total: 3–4 litres spread over a day is well within what a healthy adult on a normal diet handles, while the same amount within an hour or two is not. There is deliberately no official upper limit — the risk depends on speed and circumstances, not on one daily number [2]. And more water does not make kidneys work better: in people with chronic kidney disease, a randomised trial found that coaching to drink more did not slow the decline of kidney function [25], and for healthy adults no trial has shown that drinking beyond thirst improves how the kidneys work [9].

This is why Little Often gently notes a daily goal above 4 litres, asks you to confirm one above 6 litres, and lets you know when you log more than a litre within a single hour.

When the standard advice does not apply

Can babies drink water?

Babies under 6 months must not be given water at all — breast milk or formula covers their needs, and extra water is dangerous for their immature kidneys [26].

What changes with illness, medication or age?

With heart, kidney or liver disease, your doctor may prescribe a fluid limit — generic app goals are not for you. Some medicines lower the safe ceiling: thiazide diuretics and some antidepressants are among the most common causes of dangerously low blood sodium [27] [28]. Endurance athletes are advised to drink to thirst, not ahead of it [7]; sports-medicine bodies additionally recommend individual plans built around sweat rate, aiming to lose no more than about 2% of body weight during exercise [29]. And older adults feel thirst less sharply, so drinking on a schedule can genuinely help [1].

How much to drink with kidney stones?

If you have had a kidney stone, the advice runs the other way: drink more — enough to pass at least 2 to 2.5 litres of urine a day, which in a temperate climate usually means around 2.5–3 litres of drinks [30] [31]. This is one of the few conditions where drinking well beyond thirst has actually been tested [9]: over five years, stone formers assigned to high intake had half the recurrences of those left to their usual habits — 12% against 27% [32]. Keeping it up is the hard part: a large trial reported in 2026 coached people to drink more, raised their urine output only modestly compared with standard advice, and found no fewer stones over two years [33]. A tracker cannot make that decision for you, but it can make the habit visible.

This is not medical advice

This page is general information for healthy adults, based on the sources below. It is not a diagnosis, a treatment or a personal recommendation. If you have a chronic condition, take medication regularly, are pregnant or breastfeeding, please discuss your fluid intake with your doctor.

Sources

References are given in their published form. Hydration research is often paid for by drink manufacturers; where a study was, the entry says so. Industry funding does not by itself make a result wrong — it means the result needs independent confirmation, and the text says where that exists.