How Much Water Do You Actually Need? (And Why Thirst Isn't Too Late)

Two things this page used to tell you were wrong. One was the target; the other was the reason you supposedly need it. Both are worth correcting, because between them they have sold a great many water bottles.

How much fluid do you actually need?

Six to eight cups a day of anything drinkable — not eight glasses of plain water stacked on top of everything else you drink. The NHS guidance is that "people should aim to drink 6 to 8 cups or glasses of fluid a day", and the definition of fluid is broad: "water, lower-fat milk and sugar-free drinks, including tea and coffee, all count as part of your daily intake."

The famous "8 × 8" rule — eight eight-ounce glasses — has a thinner foundation than its fame suggests. Heinz Valtin, professor emeritus of physiology at Dartmouth and the author of two standard textbooks on the kidney and water balance, reviewed the evidence for it in the American Journal of Physiology and found "no scientific proof that absolutely every person must 'drink at least eight glasses of water a day'."

So if you have been running a daily deficit against a target nobody could source, you have been failing an exam that was never set.

Is thirst really too late?

No — and this page previously said the opposite, which is the claim most worth retracting. The usual line is that thirst means you are already dehydrated. The physiology runs the other way.

Valtin's review notes that "thirst begins when the concentration of blood has risen by less than two percent, whereas most experts would define dehydration as beginning when that concentration has risen by at least five percent."

Thirst is an early warning, firing at roughly the two percent mark — well before the five percent at which dehydration is considered to begin. It is not a late alarm. It is a sensitive one, and in healthy adults it works.

That claim is worth being angry about, mildly, because it is the load-bearing wall of an entire product category. If thirst were unreliable you would need an external system to tell you when to drink. It is not, and mostly you do not.

Do tea and coffee count?

Yes, on both the NHS's account and Valtin's. The NHS includes tea and coffee explicitly, and Valtin's review found that "caffeinated drinks should indeed count toward the daily fluid intake in the vast majority of persons."

The belief that coffee dehydrates you — that each cup somehow costs you more fluid than it delivers — does not survive the evidence. Caffeine has a mild diuretic effect; the drink is still overwhelmingly water, and the net contribution is positive.

Which means a lot of people tracking "water intake" have been undercounting their actual fluid intake by several cups a day, then feeling guilty about the gap.

So what should a hydration reminder do?

Much less than it is usually sold as doing. If thirst works, a reminder is not correcting a broken signal — it is dealing with a situation where the signal is noticed and then overridden.

That situation is real and worth a prompt:

• When drinking is inconvenient rather than un-signalled. Back-to-back meetings, a shift where you cannot leave the floor, a long drive. You register the thirst and postpone it, repeatedly.
• When something has genuinely blunted the signal. Older adults commonly experience reduced thirst sensation, and some medicines affect it. This is a real exception to everything above, and it is the case where an external prompt earns its place.
• When a clinician has actually set you a target. Some conditions come with a fluid plan — either a minimum or a restriction. Then the reminder is helping you follow a prescription rather than guessing at one.

Two or three prompts across the sedentary parts of the day is plenty. In ReminderIt you can set them by message — "remind me at 11am and 3pm to have a drink" — and it reads the schedule back before saving.

What is not worth doing is six reminders chasing a litre count, which is what this page recommended before. That is a lot of interruption in service of a number with no evidence behind it, and over-reminding a low-stakes task is how you train yourself to ignore reminders that matter.

Where this does not apply

Plenty of places, and they are the ones to take seriously.

Fluid needs rise with heat, exercise, illness involving fever, vomiting or diarrhoea, and pregnancy or breastfeeding. They are deliberately restricted in some kidney and heart conditions, where drinking to a general guideline could be actively harmful. Older adults may not feel thirst reliably. Certain medicines change the picture entirely.

None of that is something a blog can assess, and none of it is what this page is for. If you have been given a fluid target or a fluid limit, that number beats anything here and anything on a bottle.

What survives for everyone else is simple enough: the number is smaller than you were told, most drinks count toward it, and the signal you were told to distrust is the one that has been working the whole time.

Sources: NHS, Water, drinks and your health — 6 to 8 cups a day and which drinks count. Valtin H, "Drink at least eight glasses of water a day." Really? Is there scientific evidence for "8 × 8"?, American Journal of Physiology (2002), summarised by Dartmouth Medical School.