The eight-glasses-a-day rule has no study behind it. Here is what does.

There is no study behind "drink eight 8-ounce glasses of water a day." A review published in the American Journal of Physiology went looking for one in 2002 and reported that no scientific studies were found in support of the advice [s1]. The nearest thing to an evidence-based number comes from a different exercise entirely — the US Institute of Medicine set an adequate intake for total water, from drinking water, all other beverages and food combined, of 3.7 litres a day for men aged 19 to 30 and 2.7 litres for women in the same age band [s3]. That is a description of what Americans were already drinking, not a target anyone was tested against.

Where the number came from, and where it didn't

The review's author searched electronic databases, the older literature that predates them, and consulted specialists in thirst and drinking to trace the origin of the advice, shortened in the paper to "8 x 8" [s1]. What the search turned up instead was the opposite: surveys of food and fluid intake in thousands of adults of both sexes, published in peer-reviewed journals, suggesting that such large amounts are not needed, because the people surveyed were healthy [s1].

The review also dismantled the rule's usual companion clause — that coffee, tea and alcohol do not count. Published studies, it concluded, showed that caffeinated drinks, and to a lesser extent mild alcoholic beverages such as beer in moderation, may indeed be counted toward the daily total [s1].

Two limits on that conclusion are stated in the paper itself and are routinely dropped when it gets quoted. The finding applies to healthy adults in a temperate climate leading a largely sedentary life [s1]. And the author is explicit that large intakes, equal to or greater than 8 x 8, are advisable for the treatment or prevention of some diseases, and are called for in special circumstances such as vigorous work and exercise, especially in hot climates [s1].

What the official number actually is

The Institute of Medicine did not set a requirement for water in the way it sets one for iron or vitamin C. It set an adequate intake, and it set it from the median total water intake recorded in US survey data — because normal hydration can be maintained across a wide range of intakes, there was no dose to derive [s3]. The panel's own framing of how hydration is maintained is worth quoting: on a day-to-day basis, fluid intake driven by the combination of thirst and the consumption of beverages at meals allows maintenance of hydration status and total body water at normal levels [s3].

There is also no upper limit. Because healthy individuals have considerable ability to excrete excess water and thereby maintain water balance, no tolerable upper intake level was set — though the report notes that acute water toxicity has been reported [s3].

The measurement study that changed the shape of the question

In 2022, a consortium working with the International Atomic Energy Agency's doubly labelled water database published objective measurements of water turnover — the water the body actually uses in a day — in 5,604 people aged from 8 days to 96 years, using deuterium tracking rather than questionnaires [s2].

The headline result is the spread. At the low end, men and women turned over roughly 1 to 1.5 litres a day; at the upper end, around 6 litres, with outliers in the 10-litre range [s2]. The determinants were the obvious ones, quantified: athletes had about 1,000 millilitres more turnover than non-athletes with everything else held equal; turnover was about 1,000 millilitres higher at an air temperature of 30°C than at the nadir between 0°C and 10°C, and about 400 millilitres higher at −10°C than at that nadir [s2]. Turnover peaked between the ages of 20 and 30 in men and 20 and 55 in women, and was about 700 millilitres lower at age 80 than at age 30 [s2]. Across adults, water turnover tracked energy expenditure at roughly 0.33 litres per megajoule, about 1.4 millilitres per kilocalorie [s2].

The authors' own verdict on the rule is unambiguous: the common suggestion that we should drink eight 8-ounce glasses of water per day, approximately 2 litres, is not backed by objective evidence [s2].

Two things this evidence does not settle

Water turnover is not the same as required drinking water. The paper is direct about this: without measured water intakes from food, it is not possible to assess the relative contributions of food versus drinking water and beverages, and no study to date has adequately addressed the question [s2]. A person whose body turns over 3 litres a day is not thereby shown to need 3 litres from a glass.

Nor does any of this establish that drinking more than you need is beneficial. The 2002 review sets out to show the absence of evidence for a high target, not the presence of evidence for a low one [s1], and the isotope study provides prediction equations for turnover, not a prescription [s2]. The honest reading is that a healthy sedentary adult in a temperate climate has no evidential reason to count glasses, and that the numbers which do exist vary by a factor of four or more between individuals, driven by body size, activity, athletic status, pregnancy, air temperature, humidity, altitude and where in the world a person lives [s2].

A discrepancy worth recording

The published Science abstract describes the sample as drawn from 23 countries; the author manuscript deposited in PubMed Central for the same paper describes 26 [s2]. The participant count, 5,604, and the age range, 8 days to 96 years, are identical in both versions [s2]. We report both figures rather than choosing between them.

This article is informational and is not medical advice. Fluid needs differ substantially in pregnancy, in hot or high-altitude environments, during illness, and in kidney, heart and endocrine conditions, and those are questions for a clinician rather than a rule of thumb.

Sources

• "Drink at least eight glasses of water a day." Really? Is there scientific evidence for "8 x 8"? — American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 2002-11-01
• Variation in human water turnover associated with environmental and lifestyle factors — Science, 2022-11-24
• Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate — Institute of Medicine / National Academies Press, 2005