How the estimate works
The calculation starts at about 35 ml per kilogram of body weight, adds roughly 10 ml per minute of exercise, and applies a multiplier for climate. For a 70 kg person exercising 30 minutes in a warm climate: (70 × 35 + 30 × 10) × 1.1 = (2,450 + 300) × 1.1 ≈ 3.0 litres. A smaller example: a 55 kg person doing 20 minutes in a cool climate gets (1,925 + 200) × 1.0 ≈ 2.1 litres.
Be sceptical of any water formula, including this one
The 35 ml/kg figure is a widely used clinical rule of thumb, not a validated guideline, and it is worth knowing that the major reference body declined to set a target at all. The US National Academies, in their 2005 Dietary Reference Intakes, reported observed adequate intakes of about 2.7 litres of total water daily for women and 3.7 litres for men — and stressed that the vast majority of healthy people meet their needs perfectly well simply by drinking to thirst.
Two details in that matter. Those figures are total water, including the water in food, which is typically around a fifth of intake and considerably more if you eat plenty of fruit, vegetables, soup or yoghurt. And they describe what healthy people were observed to consume, not a threshold you fail by missing.
The familiar "eight glasses a day" rule has no research behind it either. It most plausibly descends from a 1945 US Food and Nutrition Board recommendation of around 2.5 litres daily — which noted in the very next sentence that most of this quantity is contained in prepared foods. The qualifier was forgotten; the number was not.
What genuinely changes your needs
- Exercise — sweat rates vary enormously between individuals, from under half a litre to over two litres an hour. The 10 ml/minute used here is a middle estimate and could easily be double or half for you.
- Heat and humidity — humidity matters because sweat that cannot evaporate does not cool you, so your body keeps producing more of it.
- Altitude — increases losses through faster breathing and increased urination.
- Illness — fever, vomiting and diarrhoea raise requirements sharply, and this is when dehydration becomes genuinely dangerous.
- Pregnancy and breastfeeding — both raise needs, breastfeeding substantially.
- Medications and conditions — diuretics increase losses, while some kidney and heart conditions require fluid restriction. If that applies to you, follow your clinician rather than this page.
The check that beats arithmetic
Urine colour is a free and reasonably reliable proxy. Pale straw suggests good hydration; dark yellow or amber suggests drinking more. Two caveats: B vitamins turn urine bright yellow regardless of hydration, and first-morning urine is normally darker.
Thirst also works better than it is given credit for. In healthy adults with access to fluid, the thirst mechanism is a competent regulator, and the widespread belief that "if you are thirsty you are already dehydrated" overstates the case considerably.
Can you drink too much?
Yes, though it is uncommon. Consuming very large volumes of plain water over a short period can dilute blood sodium — hyponatraemia — which is dangerous and occasionally fatal. It occurs mostly in endurance athletes drinking heavily over many hours. Consistently colourless urine with very frequent urination generally just means you are drinking more than you need, which is harmless but pointless.
Hitting the target
Most people who struggle with hydration are not short of motivation; they are short of a cue. Attaching a glass to something you already do reliably — waking, each meal, sitting down at your desk, before training — works better than a daily intention. A visible bottle beats one in a bag.
And everything counts. Tea, coffee, milk, and sparkling water all contribute to total fluid; caffeine's mild diuretic effect does not cancel out the water the drink contains. See how to drink more water every day for habits that stick, and the water intake chart for the full table by body weight.
Why the published figures disagree
The result above puts five figures side by side because they are not five estimates of the same thing. Each answers a different question, and once you know which, the gap between them stops looking like a contradiction.
- The National Academies figure describes; it does not prescribe. It reflects what healthy adults were typically observed to take in, food included. It is not a minimum you fall short of.
- The NHS figure counts drinks only and makes no attempt to cover water from food, which is why it is the lowest of the five.
- The body-weight rules scale in a straight line. Both 35 ml/kg and "half your weight in ounces" rise with every kilogram. That is reasonable for muscle, which is mostly water, and less so for fat, which holds far less. For a heavier person they can run well above the fixed figures.
- "Eight glasses" is a constant: the same 1.9 L for a 50 kg runner and a 100 kg desk worker. That alone shows it was never derived from anyone's physiology.
In practice, if the figures for you span less than a litre, any of them is a reasonable starting point. If they span more, usually because you are well above or below average weight, rely less on the weight-based rules and more on thirst and urine colour.