Almost every widely repeated figure about water intake is a misreading of a real reference value. It is worth being precise about what the evidence actually says.

Unproven
You should drink eight glasses of water a day.

No health authority has ever established this as a requirement, and no study supports it as a universal target. Its most likely origin is a 1945 US Food and Nutrition Board recommendation of about 2.5 litres of water daily — which explicitly added that "most of this quantity is contained in prepared foods". That second sentence was dropped, and the remainder became a rule.

  • The EFSA and NASEM reference values are for total water from all sources, not for drinking water alone
  • Requirements vary by body size, activity, ambient temperature, diet and health status by a factor of several
  • No trial has demonstrated benefit from a fixed intake in healthy adults with free access to fluid

What the reference values actually say

Adequate intake values for total water
AuthorityAdult menAdult womenBasis
EFSA (Europe)2.5 l/day2.0 l/dayTotal water from drinks and food combined
NASEM (United States)3.7 l/day2.7 l/dayTotal water from all sources

Both are adequate intakes derived from observed population intakes, not requirements derived from physiological need. Food typically contributes 20–30% of the total.

Subtract the food contribution and the drinking-fluid figure is roughly 1.5 to 2 litres for European women and about 2 to 2.5 for men — as an observed average, not a target to hit. And all fluids count: tea, coffee, milk, juice and soft drinks all contribute. The idea that caffeinated drinks are net dehydrating is not supported at normal consumption levels.

Water in the human body

Roughly 55 to 60% of an adult body is water, distributed between cells, the space around them, and blood plasma — and the proportions vary a lot between people.

Water in the human bodyTotal body water is about 55 to 60% of adult body mass. About two thirds of it is intracellular, inside cells. The remaining third is extracellular, of which most is interstitial fluid between cells and roughly a quarter is blood plasma. The proportion is higher in infants, lower in older adults, and lower in people with a higher proportion of body fat, because adipose tissue holds much less water than muscle.Total body water~55–60% of body massIntracellular fluid~2/3 of totalExtracellular fluid~1/3 of totalInterstitial fluidbetween cellsBlood plasma~1/4 of extracellularVaries by age and body composition
Why the range
Muscle is roughly 75% water; adipose tissue is roughly 10%. Body composition therefore matters more than body size.
By age
Newborns are around 75% water. The proportion falls through childhood and continues to decline in later life.
Balance, not stock
The body regulates concentration, not volume, principally through the kidneys and antidiuretic hormone. Thirst is a reliable signal for most healthy adults under normal conditions.

These are population averages. Individual values vary substantially, and clinical assessment of hydration status uses measurements rather than these proportions.

What actually regulates your intake

The body regulates the concentration of body fluids, not their volume. When plasma osmolality rises by roughly 1–2%, the hypothalamus triggers both thirst and antidiuretic hormone release, which makes the kidneys retain water. This system is precise, fast and reliable in healthy adults.

For most healthy people with free access to fluid, drinking to thirst is sufficient. Urine colour is a reasonable rough guide — pale straw suggests adequate hydration, dark yellow suggests drinking more — with the caveat that some vitamin supplements and medications colour urine independently.

When more is genuinely needed

  • Hot climates and heat waves, where sweat losses can exceed a litre per hour
  • Prolonged or intense physical activity, especially in heat
  • Fever, vomiting or diarrhoea
  • Pregnancy and breastfeeding, which add roughly 300 ml and 700 ml per day respectively
  • Certain medical conditions and medications — follow clinical advice rather than general guidance

Overhydration is real

Does tea or coffee count?

Yes. At normal consumption levels the diuretic effect of caffeine does not offset the fluid provided. Both contribute net water.

Is drinking more water good for my skin?

In people who are adequately hydrated, there is little evidence that additional intake improves skin appearance. Dehydration does affect skin, but the fix is correcting dehydration, not exceeding normal intake.

Does drinking water help with weight loss?

Replacing sugary drinks with water reduces calorie intake, which helps. Water itself has no metabolic effect on weight beyond a small, short-lived thermogenic response.

Is it bad to drink water with meals?

No. The idea that it dilutes digestive enzymes is not supported by physiology.

Should I drink more if my urine is dark?

It is a reasonable prompt to drink more, with the caveat that B vitamins and some medications darken urine independently of hydration.

Sources

More on water & the human body