Hydration: what the evidence actually says about '8 glasses a day'

Where "8 glasses a day" came from

The advice to drink eight 8-ounce glasses of water daily is one of the most repeated pieces of health guidance in circulation, and it doesn't have a clear scientific origin. Heinz Valtin, professor emeritus of physiology at Dartmouth Medical School, searched for the evidence behind it in a 2002 invited review in the American Journal of Physiology and found no supporting studies at all (full text). His best explanation for where the number came from is a 1945 U.S. Food and Nutrition Board recommendation of roughly 2–2.5 quarts of water per day — a figure that explicitly included water from food, not just drinking water. That caveat appears to have been dropped somewhere in the decades of repetition, turning "about 2.5 quarts total, mostly from food and drink combined" into "drink eight glasses," a materially different and unsupported claim.

What the reference bodies actually recommend

Two major bodies have published formal adequate-intake values for water, and they don't fully agree with each other, which is itself informative about how much precision this kind of estimate can really claim. The European Food Safety Authority's 2010 scientific opinion sets an adequate total water intake — from all beverages and food combined — of 2.0 L/day for women and 2.5 L/day for men, under conditions of moderate temperature and moderate activity (EFSA Journal 2010;8(3):1459). The U.S. Institute of Medicine's 2004 Dietary Reference Intakes set a higher adequate intake of total water — 2.7 L/day for women and 3.7 L/day for men — based on average intakes observed in a U.S. national nutrition survey rather than the European survey data EFSA used. Both figures are "adequate intake" estimates, not strict minimums or requirements, and both explicitly include water obtained from food, not just water you drink.

Worked example: reconciling the numbers with the calculator

The water intake calculator on this site estimates drinking water specifically, using a common sports-nutrition baseline of about 33 ml per kilogram of bodyweight. For a 75 kg sedentary adult in a temperate climate, that comes out to 75 × 0.033 ≈ 2.5 L/day — which lines up closely with EFSA's 2.5 L/day total-water figure for men, even though the two numbers are measuring slightly different things (drinking water only, versus drinking water plus food moisture). The IOM's higher 3.7 L/day figure for men looks like a bigger gap by comparison, but a meaningful share of that total is expected to come from food rather than drinking water alone — the two organizations aren't as far apart on drinking-water volume specifically as the headline totals suggest.

Adjusting for exercise and heat

Sweat losses during exercise vary widely between individuals and conditions, which is why the American College of Sports Medicine's position stand on exercise and fluid replacement, led by Sawka and colleagues, recommends fluid replacement be matched to an individual's own sweat losses rather than a single fixed number for everyone, with needs rising accordingly in heat and during prolonged or intense exercise (PubMed). The calculator on this site applies a simpler planning estimate — roughly 350 ml added per 30 minutes of exercise, plus 500 ml for a hot climate — as a reasonable starting point rather than an individualized sweat-rate calculation.

Worked example: exercise and heat combined

An 80 kg person doing 60 minutes of exercise in a hot climate: base intake 80 × 0.033 ≈ 2.6 L, plus 0.35 L × (60 ÷ 30) = 0.7 L for exercise, plus 0.5 L for heat, totaling 3.8 L/day — roughly 1.3 L more than the same person's sedentary baseline. That additional 1.3 L reflects the two biggest drivers of extra fluid need (exercise duration and ambient heat), consistent with the direction — though not the individualized precision — of the ACSM guidance above.

The other side: overhydration and hyponatremia

More water is not simply better past a certain point. Exercise-associated hyponatremia — abnormally low blood sodium caused by drinking more fluid than the body loses, diluting sodium concentration — is a documented risk in endurance events. Almond and colleagues measured blood sodium in participants of the 2002 Boston Marathon and found 13% had hyponatremia (serum sodium ≤135 mmol/L) at the finish, with 0.6% experiencing critical hyponatremia (≤120 mmol/L), published in the New England Journal of Medicine (PubMed). The strongest risk factor identified was weight gain during the race — direct evidence of drinking more than was being lost through sweat — along with race times over four hours, meaning slower runners drinking steadily over a longer period were at higher risk than faster ones. This isn't an argument against hydrating during exercise; it's evidence that matching intake to actual losses, rather than drinking as much as possible "to be safe," is the safer strategy in both directions.

Practical signals that don't require a formula

Thirst, while imperfect during intense short-duration exercise, is a reasonable day-to-day guide for most people outside of prolonged endurance events. Urine color is a simpler practical check: pale yellow generally indicates adequate hydration, while consistently dark urine suggests under-hydration and consistently clear urine (especially combined with unusual swelling or confusion during long exercise) can be a sign of overhydration risk rather than a goal to chase.

Using the calculator sensibly

Treat the water intake calculator's output as a general planning estimate for drinking water specifically, not a hard target that overrides thirst, and be especially cautious about deliberately overshooting it during long endurance events, where the marathon research above shows real harm from excess intake. Adjust upward for genuine heat or heavy training days, and let day-to-day thirst and urine color fine-tune the number rather than forcing an exact liter count every day.

---

This article summarizes published hydration research and is general information, not medical advice. It doesn't account for kidney conditions, medications, pregnancy or other factors that change individual fluid needs — talk to a doctor about your own hydration targets if you have a relevant medical condition.