How much water to drink a day: why there's no single figure for everyone

AI-generated content. How we make it

In brief

  • For adults, EFSA sets an adequate intake of 2.0 liters (about 68 fl oz) of water a day for women and 2.5 liters (about 85 fl oz) for men. This is total water: from all beverages and from food, under conditions of moderate temperature and physical activity.1
  • Food usually provides around 20 to 30% of that water, so what you drink is less than the total figure; how much less depends on what you eat.1
  • The “eight glasses a day” don't come from any study: a 2002 review found no evidence to back them, and a 2022 study of 5,604 people concluded that one figure for everyone doesn't work.34
  • Heat, physical activity, pregnancy and breastfeeding increase what you need; with age, thirst gives less warning. And drinking much more than you lose also carries risks, especially in long endurance events.169
  • If an illness or a medication changes what you should drink (for example, if you've been told to limit fluids), what counts is what your health professional tells you, not these population figures.1

“You should drink two liters of water a day” and “eight glasses a day” are two of the most repeated pieces of health advice. There's some truth in them and quite a lot of simplification: the European reference isn't about the water you drink but about total water, and it doesn't apply in the same way to a winter day at the office and to an August afternoon on a bike.

This article explains exactly what the references say, where the round numbers come from, what raises or lowers what you need, what you can expect from thirst and urine color, and why drinking too much isn't a good idea either. It's part of the guide to healthy eating, which reviews the dietary pattern that what you drink fits into.

What the European reference says

In 2010, the nutrition panel of the European Food Safety Authority (EFSA) published its reference values for water. Because what each person needs varies widely with the individual and the environment, it didn't set an exact requirement but adequate intakes: amounts it calculated from what groups of healthy people drink and eat and from how much the kidneys should concentrate urine to leave a safety margin.1

Three details change how to read the figures: they're for total water (tap or bottled water, beverages of all kinds and the water in food), they only hold with moderate temperature and physical activity, and they're references for groups of healthy people, not what a specific person needs.1 Spain's food safety agency, AESAN, includes the same figures in its recommendations for the Spanish population.2

Figure 1. Adequate intake of total water per day, according to EFSA

Adequate intakes of total water (beverages of all kinds plus the water in food) set by EFSA in 2010 for the healthy European population, under moderate ambient temperature and moderate physical activity. The pregnancy and breastfeeding bars add the extra 300 ml and about 700 ml proposed in the opinion to the reference for women. Children under 2 aren't shown; for them the opinion gives ranges or figures per kilogram of body weight.1 AESAN gives the same figures by age and sex for Spain.2 These are population references, not the amount a specific person needs.

Where the eight glasses and the two liters come from

In 2002, the physiologist Heinz Valtin reviewed the literature and consulted specialists in thirst and fluid intake to trace the origin of the advice to drink “at least eight glasses of water a day,” which, with 8-ounce glasses, adds up to about 1.9 liters. He found no scientific study to back it.3

He did find two leads. One points to a 1974 nutrition book whose authors wrote, in passing and without citing evidence, that an average adult needs around 6 to 8 glasses a day, which could be coffee, tea, milk, soft drinks or beer, and that fruit and vegetables also provide water. The other, which Valtin takes from other authors, leads back to 1945, when the US Food and Nutrition Board proposed 2.5 liters a day for adults and added that most of that amount was already contained in prepared foods. According to that explanation, this last sentence was forgotten and the figure came to be read as water you have to drink.3

Valtin's conclusion is limited to healthy adults in a temperate climate who move little, and he notes that hard exertion or heat calls for more. Conversely, he warns that urging everyone to drink that much could be counterproductive: in specific cases, because of the risk of hyponatremia we discuss below, and in general because it makes many people feel guilty for falling short.3 The UK's National Health Service still talks about 6 to 8 glasses or cups of fluid a day, but it counts tea, coffee and milk, and presents this as guidance that changes with heat, prolonged physical activity, pregnancy or breastfeeding.5

A study published in 2022 used isotope-labeled water to measure how much water enters and leaves the body each day (water turnover) in 5,604 people aged 8 days to 96 years. Turnover was associated with body size and composition, age, physical activity, pregnancy, the country's level of development and climate. The authors conclude that a single figure doesn't work for everyone and that the eight glasses have no objective support.4

The water that comes with food and with other beverages

EFSA assumes that foods normally provide between 20 and 30% of total water and beverages between 70 and 80%, and warns that this split isn't fixed: it depends on what you eat and drink. Fruit and vegetables are more than 80% water; hot dishes, between 40 and 70%; milk, about 90%; and bakery products, less than 40%. A diet rich in fruit and vegetables provides quite a lot of water; fast food, little.1

If you do the math with those percentages, the 2.0-liter reference works out to something like 1.4 to 1.6 liters of beverages a day, and the 2.5-liter one to 1.75 to 2 liters. That's our own rough calculation, not an EFSA figure, and it changes as soon as what you eat changes.

In Spain, AESAN recommends making water your main beverage, from the tap whenever possible, and points out that coffee, tea, herbal teas and unsweetened sparkling water also help keep you hydrated, as do fruit, vegetables, gazpacho, soups and purées.2 To compare the sugar in packaged drinks, it helps to know how to read the nutrition label.

On coffee, EFSA concludes that, under normal conditions, moderate habitual caffeine intake doesn't impair hydration; only for people who take high amounts of caffeine (around 600 mg) or alcohol (around 50 g) every day might it make sense to discount some of what they drink.1 Coffee counting as fluid doesn't mean the time you have it doesn't matter: we explain separately why there's no cutoff time that works for everyone. Alcohol, like caffeine at high doses, increases urine output in the hours after you drink it,1 and on top of that it doesn't help you rest better at night.

What raises how much you need

Heat and physical activity. When you move, your body produces heat and gets rid of it by sweating. According to EFSA, sweat can reach 1 or 2 liters an hour depending on the intensity and duration of the effort, clothing, air and humidity, and exceed 1.5 liters an hour during long exertion in the heat. Under extreme conditions of temperature and exercise, losses can reach about 8 liters a day, and then the salts lost in sweat also need replacing.1

During heat waves, Spain's Ministerio de Sanidad advises drinking water or other fluids often even if you're not thirsty; avoiding drinks with caffeine or alcohol or with a lot of sugar, because they can promote dehydration; and consulting a health professional if symptoms appear that last more than an hour and may be due to the heat. It asks for special attention to babies and children, pregnant women, breastfeeding women, older people and people with illnesses that heat can make worse.10 If you've been told to limit fluids because of an illness, this general recommendation doesn't replace what your health professional tells you: ask them what to do on hot days. On caffeine, this advice is more cautious than the EFSA conclusion for normal conditions we saw earlier: they're recommendations for different situations, and in a heat wave the guidance to follow is the Ministerio's. This comparison is ours.

Pregnancy and breastfeeding. EFSA adds 300 ml a day during pregnancy and about 700 ml while breastfeeding, to cover the water that goes into the milk.1

Age. In a classic experiment, seven healthy, active men aged 67 to 75 and seven aged 20 to 31 went 24 hours without drinking. Afterward, the older men's blood had become more concentrated, but they felt less thirsty and drank less, and they didn't get back to their starting point.6 It's a very small study, and only in men, but it points in the same direction as what EFSA reports: older people are more at risk of drinking too little because thirst and appetite decline and the kidneys concentrate urine less well. In a study it cites, of 1,372 older people living independently, total water fell short of the amount recommended at the time in 28% of the youngest in the group and in 41% of the oldest.1 That's why EFSA kept the same reference for older people as for other adults, even though they eat less.1 AESAN adds that these changes weigh more heavily when there's cognitive or functional decline.2

Thirst and urine color: useful, with limits

The body regulates its water quite precisely, which is why you can be well hydrated with very different amounts of water.1 According to EFSA, losing 2 or 3 liters of sweat over a few hours of activity in the heat is usually made up over the next 24 hours because people drink more.1 Valtin relies on that precision, and on surveys of thousands of healthy adults who drank less than eight glasses a day, to conclude that healthy adults in a temperate climate don't need to force themselves to drink them.3

Urine color is another clue. The NHS suggests drinking enough for your urine to be pale yellow.5 EFSA considers it a useful but imprecise indicator: color doesn't track hydration status exactly and changes with diet and with some medications.1

With age, neither signal is enough on its own. A Cochrane review brought together 24 studies of 3,412 people aged 65 or older, living at home or in care homes or admitted to the hospital, and analyzed 67 simple tests. Neither urine color, nor feeling thirsty, nor noticing a dry mouth worked on its own to tell whether an older person was drinking enough, and the authors advise against using them that way.7

Drinking too much isn't the answer either

An adult's kidneys can excrete at most around 0.7 to 1 liter of water an hour. If for hours you drink more than the kidneys can excrete, the sodium in the blood becomes diluted (hyponatremia), and that can become serious. According to EFSA, it's unlikely under normal conditions and with normal habits, but it can happen during or after very long endurance efforts, when the kidneys also tend to excrete less water than usual because of a hormone that retains it.19

In the 2002 Boston Marathon, of 488 runners who gave a blood sample at the finish line, 13% had low sodium and 0.6% very low sodium. In the analysis that accounted for several factors at once, it was associated with having gained weight during the race, a finishing time of more than four hours and a very high or very low body mass index; looking at one factor at a time, also with having drunk more than 3 liters during the race.8 The 2015 international consensus on this exercise-associated hyponatremia considers weight gain during such an event a sign that people drank more than they were losing, and concludes that the main cause is drinking beyond what's lost through sweat and urine, that in these events the safest approach for most people is to drink when thirsty, and that the advice to drink before you're thirsty has encouraged overdrinking.9

This may seem to clash with the heat-wave advice to “drink even if you're not thirsty.” Read carefully, they're not talking about the same thing: the consensus warns against drinking more than you lose during an endurance event,9 and the Ministerio de Sanidad asks people, on days of extreme heat, to drink often even if they're not thirsty, whether or not they're physically active, and to avoid outdoor sports in the middle of the day; it addresses the whole population, including older people, in whom thirst gives less warning.106 Drinking often when you sweat a lot isn't the same as drinking beyond what you lose. This combined reading is ours, not the sources'.

How to use these figures in your daily life

A few practical ideas can be drawn from all of the above. They're our own deductions from the sources, not findings of a study:

  • Take the 2.0 and 2.5 liters as a reference for total water on normal days, not as a target for the water you drink, and count herbal teas, coffee, milk, soups, fruit and vegetables too.
  • In the heat or with a lot of activity you'll need more: keep water at hand instead of setting yourself another number.
  • If you're well on in years or look after an older person, and no one has told you (or the person you look after) to limit fluids, don't rely on thirst alone: drinking at set times, such as with meals, doesn't depend on a signal that has grown weaker.
  • In a long endurance event, don't force yourself to drink beyond your thirst.
  • If an illness or a treatment changes what you should drink, ask your health professional.
There's no water figure that works for everyone and for every day. The European reference, 2.0 liters for women and 2.5 for men, includes the water in food and applies to moderate conditions. From there, heat, physical activity, pregnancy, breastfeeding and age change what you need, and if an illness or a medication comes into play, your health professional is the one who decides.

Keep reading

Want to turn information into action?

naro helps you prioritize 2–3 high-potential habits and sustain them over time, with context and follow-up.

Sources

Each small number in the text leads to the source that statement rests on. Practical advice without a number is a deduction from these sources, and examples with a person's name are made up. It is general information: it does not replace an assessment by a health professional. How we make this content

Sources consulted up to 28 October 2026.

  1. Institutional scientific opinion. EFSA Panel on Dietetic Products, Nutrition, and Allergies (NDA) (2010). Scientific Opinion on Dietary Reference Values for water. EFSA Journal. Adequate intakes of total water for the healthy European population, under moderate temperature and physical activity. Based on observed intakes and on a desirable urine concentration, not on trials.
  2. Institutional report. Comité Científico de la AESAN (2022). Informe del Comité Científico de la AESAN sobre recomendaciones dietéticas sostenibles y recomendaciones de actividad física para la población española. Revista del Comité Científico de la AESAN. Adopts EFSA's adequate intakes for Spain and recommends water as the main beverage.
  3. Narrative review. Valtin H (2002). "Drink at least eight glasses of water a day." Really? Is there scientific evidence for "8 × 8"?. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology. Traces the origin of the advice and the evidence supporting it. Its conclusions are limited to healthy, sedentary adults in a temperate climate.
  4. Observational study with isotope measurement. Yamada Y, Zhang X, Henderson MET, et al. (2022). Variation in human water turnover associated with environmental and lifestyle factors. Science. 5,604 people aged 8 days to 96 years. Measures water turnover, not the amount you need to drink.
  5. Institutional information for the public. NHS (UK National Health Service) (2023). Water, drinks and your health. NHS. Page reviewed on May 17, 2023.
  6. Experimental study. Phillips PA, Rolls BJ, Ledingham JG, et al. (1984). Reduced thirst after water deprivation in healthy elderly men. New England Journal of Medicine. Seven older and seven young men, 24 hours without drinking. Very small and only in men. We could only read the abstract.
  7. Cochrane systematic review. Hooper L, Abdelhamid A, Attreed NJ, et al. (2015). Clinical symptoms, signs and tests for identification of impending and current water-loss dehydration in older people. Cochrane Database of Systematic Reviews. 24 studies of 3,412 people aged 65 or older, without kidney or heart failure.
  8. Cohort study. Almond CS, Shin AY, Fortescue EB, et al. (2005). Hyponatremia among runners in the Boston Marathon. New England Journal of Medicine. 488 non-elite runners in the 2002 Boston Marathon who gave a blood sample at the finish line. We could only read the abstract.
  9. Consensus statement. Hew-Butler T, Rosner MH, Fowkes-Godek S, et al. (2015). Statement of the 3rd International Exercise-Associated Hyponatremia Consensus Development Conference, Carlsbad, California, 2015. British Journal of Sports Medicine. Concerns physical activity and sporting events, not daily life.
  10. Institutional recommendations. Ministerio de Sanidad (2024). Sanidad recuerda a la población la importancia de tomar medidas preventivas para protegerse de las altas temperaturas. Ministerio de Sanidad (press release, July 19, 2024). Ten-point list of recommendations from Spain's Plan Nacional de Actuaciones Preventivas de los Efectos del Exceso de Temperaturas.

← The whole library