A real recommendation, missing one sentence
"Drink eight glasses of water a day" is the rare health myth with a genuine document behind it. It is usually traced to the Food and Nutrition Board of the US National Research Council, which in 1945 published guidance on how much water people need.
The interesting part is not that the advice was invented. It is that the popular version manages to be wrong in two opposite directions at the same time — it asks for noticeably less water than the 1945 figure, while being far stricter than the original about where that water has to come from. One dropped sentence did both.
What the 1945 text actually said
The recommendation had three parts, and only the first survives in popular memory.
An amount. A suitable allowance for adults was given as about 2.5 litres a day in most cases.
A way of scaling it. The board offered a rule tying water to food intake — roughly one millilitre of water for each calorie eaten. On a 2,000-calorie diet that comes to 2 litres, which is where the other familiar figure comes from. It is a sensible way to scale: bigger people eating more food need more water, and the calorie count is a rough proxy for body size and activity at once.
And the sentence that got lost. The text noted that most of that quantity is already contained in prepared foods.
That last clause is the whole argument. The 2.5 litres was never a drinking target — it was a total, of which the greater part was expected to arrive as part of ordinary meals. Strip the sentence and you convert an estimate of total water throughput into a daily drinking quota, which is a completely different instruction.
The arithmetic does not even match
Here is the part that rarely gets checked. Eight glasses of eight fluid ounces is 64 fl oz, which is 1.89 litres.
| Amount | Comes from | |
|---|---|---|
| 1945 recommendation | 2.5 L | Food and drink together, mostly food |
| "Eight glasses" rule | 1.89 L | Plain water only |
The myth asks for about 24% less total water than the document it claims to be based on — and then insists the entire reduced amount be drunk as plain water, when the original said most of it would come from food. To hit 2.5 litres in eight-ounce glasses you would need about ten and a half of them, and only if you ate nothing at all.
So the rule is not a slight simplification. It is a different claim wearing the authority of an old one.
How the sentence got dropped
Nobody can point to the moment it happened, and it is worth being honest that the trail is partly reconstruction. The most careful account is a 2002 review by the physiologist Heinz Valtin, who went looking for the evidence base behind the eight-glasses advice and could not find one — no study establishing that healthy adults in temperate climates need that much plain water, and no obvious origin except the 1945 text with its caveat removed.
The mechanics of the spread are ordinary enough. "Eight glasses a day" is short, countable and easy to repeat, while "about 2.5 litres of total water, most of which is in your food" is none of those things. Advice that fits on a fridge magnet outcompetes advice that needs a paragraph, and it was repeated by magazines, by wellness writers and, for a long time, by clinicians who had learned the rule rather than the source.
It is often suggested that the bottled water industry gave the rule a push once it became commercially convenient. That is plausible and frequently asserted, but it is not documented in the way the 1945 text is, and it is worth separating the two.
What is clear is that the rule got stricter as it travelled. The modern folk version has added claims the 1945 board never made — that only plain water counts, that coffee and tea work against you, that thirst is an unreliable signal and you should drink ahead of it.
What the guidance says now
Current figures are, if anything, further from the myth than the 1945 ones. The US National Academies suggest roughly 3.7 litres of total water a day for men and 2.7 litres for women — noticeably more than 2.5 litres, and explicitly counting everything: plain water, other drinks, and the water in food.
Two consequences follow, and they point in opposite directions from the myth.
Other drinks count. Tea, coffee, milk, juice and soft drinks are mostly water and the body uses that water. Caffeine has a mild diuretic effect, but at ordinary intakes it does not offset the fluid the drink delivers — a cup of coffee leaves you ahead, not behind. There may be excellent reasons to prefer water over a sugary drink, but hydration arithmetic is not one of them.
Food counts, and more than people expect. Many vegetables and fruits are over 90% water. Soups, stews, yoghurt, cooked grains and pasta all carry substantial water. This is exactly what the 1945 sentence was saying, and it is why the totals above look so much larger than any drinking target you have been given.
So what should you actually do
Less than the rule implies, and with less bookkeeping.
- Drink when you are thirsty. For healthy adults going about an ordinary day, thirst tracks fluid needs perfectly well. The idea that thirst arrives too late to be useful is part of the folk version, not the guidance.
- Count everything. Every drink counts, and so does what you eat. If you are tallying at all, tallying only plain water will systematically understate where you are.
- Adjust for the day, not for a target. Heat, sustained physical work and long exercise genuinely raise what you need, sometimes considerably. A cool day at a desk does not.
- Urine colour is a rough, free check. Pale is generally a sign things are fine; consistently dark suggests drinking more. It is a nudge, not a measurement.
- More is not automatically better. Deliberately drinking far beyond thirst — the gallon-a-day style challenge — is not a free health upgrade, and at the extremes it stops being harmless. There is no benefit waiting above the amount your body is asking for.
If you want a number to start from rather than a rule to obey, our water intake calculator works from body weight and activity and shows the reasoning, including how much of the total is expected to come from food.
What this cannot tell you
- No figure here is a personal target. 2.5, 2.7 and 3.7 litres are population averages. Body size, climate, activity and what you eat move the real number substantially in both directions.
- The water in food cannot be counted precisely. It depends entirely on what you eat — a day of soup and fruit and a day of bread and cheese are not close. Any split between "from food" and "from drink" is an estimate.
- The history is partly reconstruction. The 1945 text is documented; the exact path by which its caveat was lost is inferred, and anyone claiming to know precisely how the myth started is going beyond the record.
- This is about ordinary days for healthy adults. Endurance events, heavy heat exposure and work in hot conditions have their own requirements that a general figure does not cover.
- Fluid intake is not a lever for anything else. Claims that hitting a water target clears toxins, transforms skin or drives weight loss are not what any of the guidance above says.
Frequently asked questions
Is the eight glasses rule completely made up?
Not made up, but badly garbled. It descends from a real 1945 recommendation of about 2.5 litres of total water daily, most of which the text expected to come from food. The popular version drops the food clause and lands on 1.89 litres of plain water — both less total water than the original and a much stricter demand about its source.
Does coffee count towards hydration?
Yes. Caffeine is mildly diuretic, but at normal intakes that effect does not cancel out the water in the cup, so a coffee leaves you net ahead on fluid. The belief that caffeinated drinks are dehydrating is one of the accretions the rule picked up along the way rather than anything the guidance says.
How much water should I actually drink each day?
There is no single correct number, which is precisely why the memorable rule spread. Current guidance puts total water — food included — at roughly 3.7 litres a day for men and 2.7 for women on average, so the drinking portion is a good deal less than either figure. For most people on an ordinary day, drinking to thirst and eating normally handles it without counting.
Can you drink too much water?
Yes. Drinking far beyond thirst, particularly large volumes quickly during prolonged exercise, is not harmless, and it is one of the reasons a fixed daily quota is a poor substitute for listening to your body. The practical version is simple: there is no prize above the amount you actually need, so the target-chasing is effort spent for nothing.
For a starting figure based on your own weight and activity, with the method shown rather than a single number asserted, see the water intake calculator, or the rest of food and weight tools.
HealthGuideTip provides general health information for educational purposes only and is not medical advice. Always consult a qualified healthcare professional about your own health, and before changing any treatment, medication, or lifestyle.