Does Coffee Dehydrate You? Not at Normal Doses, and the Trials Are Fairly Clear

Short answer: no, not at the amounts most people drink. In a crossover trial of 50 men who were already regular coffee drinkers, four 200 mL cups a day (about 308 mg of caffeine) produced the same total body water and the same 24-hour urine volume as the same amount of plain water. A separate trial of 13 drinks found that coffee, tea and cola were retained no differently from still water over four hours. Caffeine does have a diuretic effect, but reviews place the threshold at roughly 250–300 mg in one dose, in people who have gone without caffeine for days or weeks. A meta-analysis put the average extra urine at about 109 mL, less than the volume of the drink that delivered it. The evidence is mostly from men and from short studies.

What the main studies found

Is coffee a diuretic?

Caffeine is, mildly. That part of the folk belief is correct. A diuretic is something that increases urine production, and caffeine in a large enough single dose does that in people who are not used to it.

The 2003 review by Maughan and Griffin, which searched the literature back to 1966, put the dose at 250–300 mg or more, which the authors equated to about two to three cups of coffee or five to eight cups of tea taken at once. They added two qualifications that tend to be dropped when the finding is repeated. First, the effect was seen in people who had been deprived of caffeine for days or weeks before the test. Second, it was short-lived. In people who drink tea or coffee regularly, the review found a "profound tolerance", with the diuretic action much reduced. Amounts of caffeine equivalent to a normal serving of tea, coffee or a soft drink appeared to have no diuretic action at all.

The 2015 meta-analysis by Zhang and colleagues gives the size of the effect. Pooling 16 studies with a median caffeine dose of 300 mg, urine volume rose by an average of 109 mL, or 16%, compared with no caffeine. The spread around that average was wide (the standard deviation was 195 mL), so the size of the effect varied a good deal from one comparison to the next. The overall effect size was small, 0.29.

For scale, 109 mL is under four fluid ounces. A dose of 300 mg of caffeine taken as coffee comes with far more liquid than that. This is the central point: a diuretic drink only dehydrates if it makes the body lose more water than the drink contains. At ordinary coffee strength, the measured extra loss is a fraction of the cup.

The trial that measured body water directly

Most older studies on this question gave caffeine as a capsule to people who had abstained beforehand, and then measured urine for a few hours. That design is good at detecting a diuretic effect and poor at answering the question people actually ask, which is whether drinking coffee every day leaves them short of water.

The 2014 trial by Killer, Blannin and Jeukendrup at the University of Birmingham was built to answer the everyday question. The details matter:

• Participants: 52 healthy, non-smoking men aged 18 to 46 were recruited and 50 completed both arms. All were habitual coffee drinkers, taking 3 to 6 cups a day.
• Design: a counterbalanced crossover. Each man did both conditions, in random order, separated by a 10-day washout.
• Coffee arm: four 200 mL mugs of black instant coffee a day, dosed to give 4 mg of caffeine per kilogram of body mass. Mean intake worked out at 308 mg of caffeine a day, ranging from about 204 to 453 mg.
• Water arm: four 200 mL servings of tap water in place of the coffee.
• Main measure: total body water by deuterium oxide dilution, a tracer method that is considerably more direct than urine colour or body weight.

The result was a null in every measure that bears on hydration. Total body water was 51.5 kg on coffee and 51.4 kg on water (p = 0.90). Twenty-four-hour urine volume was 2,409 mL on coffee and 2,428 mL on water. Urine osmolality, urine specific gravity, haematocrit, serum sodium, serum osmolality and body mass did not differ between conditions.

Coffee against other drinks: the hydration index

The second useful experiment asked a broader question. Maughan and colleagues gave 72 men one litre of a drink over 30 minutes, then collected all their urine for four hours. Each man drank still water and three of twelve other drinks on separate occasions. The amount retained relative to water was expressed as a beverage hydration index, with still water set at 1.0.

Three things are worth taking from this table. A full litre of coffee, drunk in half an hour by people who had fasted, did not produce more urine than a litre of water. The drinks that beat water were the ones carrying sodium, potassium, protein or energy, which slow the passage of fluid and help the body hold onto it. And the authors' own summary of the caffeine question matches the 2003 review: an acute dose below about 250–300 mg is unlikely to have a measurable effect on urine output, and the 96 to 212 mg delivered by their caffeinated drinks did not.

The practical reading is that the composition of a drink matters more than whether it contains caffeine. That is also why the fastest route to rehydrate after real fluid loss involves salt and not simply more water.

Where the myth came from

The idea is old. Killer and colleagues cite a 1928 paper by Eddy and Downs as the first scientific report of caffeine’s diuretic effect. They also note something that the popular version leaves out: the same early work suggested that regular use leads to tolerance. The qualification was there from the start; it was the headline finding that travelled.

How much coffee is dehydrating?

There is no dose of ordinary coffee in these studies that produced net dehydration. The more answerable question is the dose at which a diuretic effect becomes detectable, and for whom.

To translate the threshold into cups: the FDA lists regular brewed coffee at 113 to 247 mg per 12 fluid ounces, a very wide range that depends on the bean and the brew. On those figures, 250–300 mg is somewhere between one strong 12-ounce cup and nearly three weak ones, taken in one sitting, by someone who does not usually drink it. That person will probably notice an extra trip to the bathroom. They will also have drunk between 12 and 36 ounces of water in the process.

The FDA's general figure for healthy adults is that 400 mg a day is an amount not generally associated with negative effects, which it describes as about two to three 12-ounce cups. That limit exists for reasons other than hydration: sleep, heart rate, anxiety. How long a given dose stays active is a separate matter, covered in the article on how long caffeine lasts.

Tolerance, sex and exercise: who the average hides

Tolerance. This is the largest modifier and it was noticed in the earliest work. People who take caffeine daily show a much smaller diuretic response than people who have abstained. The Killer trial recruited only habitual drinkers, which is the right population for the everyday question and the wrong one for predicting what happens to someone who has a rare triple-shot drink. People who have recently stopped and restarted are, for this purpose, non-habituated again; the timeline of that reset overlaps with caffeine withdrawal.

Sex. The Zhang meta-analysis found the diuretic effect was considerably larger in women (effect size 0.75) than in men (0.13, which was not statistically distinguishable from zero). This is a subgroup result from a modest number of studies and should be held loosely, but it points directly at the gap in the two best trials, both of which enrolled only men.

Exercise. In the same meta-analysis, caffeine taken at rest had a clear diuretic effect (0.54) and caffeine taken around exercise did not (0.10, not significant). The authors' conclusion was that concerns about unwanted fluid loss from caffeine are unwarranted, particularly when it is taken before or during exercise. A coffee before a run is not a hydration problem on this evidence.

What the studies cannot tell you

A sceptical reading has to include the limits, and there are several.

The participants were men. All 50 in the body-water trial and all 72 in the hydration index trial. Given that the pooled data suggest women respond more strongly, the reassuring result cannot simply be copied across.

They were short. The Killer trial ran each condition for a few days; the paper's abstract says three and its methods section says four, an inconsistency in the published text. The hydration index trial lasted four hours per drink. Neither is a long-term study, though there is no obvious reason tolerance would reverse over months of continued use.

The dose was moderate. About 300 mg a day, split into four servings. The trial says nothing about 600 or 800 mg a day, or about one very large dose.

There was no decaf arm. The authors note that without one they could not separate caffeine from the other compounds in coffee, although with a null result there was little to separate.

The funding is from interested parties. The Killer trial was funded by the Institute for Scientific Information on Coffee, and one author was employed by PepsiCo at the time of publication. The hydration index trial was funded by the European Hydration Institute, whose scientific advisory board was chaired by the lead author. The papers state that funders had no role in design or analysis, and the findings line up with the independent 2003 review and the 2015 meta-analysis. It is still something a reader should know.

Nobody studied people who were already dehydrated, ill, elderly or pregnant. The trials used healthy, rested, normally hydrated adults.

Does coffee count as water intake?

According to the body that sets the US reference values, yes. The National Academies' Dietary Reference Intakes set an Adequate Intake for total water of 3.7 litres a day for men and 2.7 litres for women, a figure that includes the water in food and in every drink. About 80% of that comes from drinking water and other beverages, which the report puts at roughly 13 cups a day for young men and 9 for young women. On caffeine, the report’s summary says that although some studies show a diuretic effect from caffeinated drinks, the effect may be transient, and such beverages contribute to total water intake. The news release that accompanied the report puts it more plainly: caffeinated beverages can contribute as much as non-caffeinated options.

The same report did not recommend a fixed number of glasses. It concluded that most healthy people meet their needs by letting thirst guide them. The "eight glasses of water, and coffee does not count" rule has no basis in the reference document it is usually assumed to come from.

What is not worth doing

Matching every coffee with a glass of water "to cancel it out". In the only trial that measured total body water, 800 mL of coffee a day and 800 mL of water a day gave the same result. There is nothing to cancel. Drinking water alongside coffee is harmless and can be pleasant; it is not a correction.

Subtracting coffee from a daily fluid tally. Hydration apps and bottle-tracking routines that count coffee as zero, or as negative, are working against the reference intake they are trying to meet.

Assuming coffee is therefore the best thing to drink in every circumstance. It is not. After heavy sweating, vomiting or diarrhoea, drinks with sodium are retained better than water or coffee. In a dry aircraft cabin the question is how much fluid of any kind is going in, as the piece on whether flying dehydrates you sets out.

What to actually do

Count it. For a healthy adult who drinks coffee regularly, the cup counts toward total water much as water does. Tea counts too.

Keep an eye on the caffeine total for other reasons. The FDA's 400 mg a day is a reasonable ceiling for most adults. People who are pregnant, trying to become pregnant or breastfeeding are advised by the FDA to ask their health care provider about limits. The reasons to moderate coffee are its effects on sleep and nerves, and its benefits are on the side of alertness, which is examined in the article on whether coffee helps you focus.

Questions people ask

Does coffee dehydrate you? Not at moderate intake in regular drinkers. In a crossover trial of 50 men, four 200 mL cups a day, about 308 mg of caffeine, gave the same total body water and 24-hour urine volume as an equal amount of water. The evidence comes mainly from men and from studies lasting days.

Is coffee a diuretic? Caffeine is a mild one. Reviews place the effect at single doses of roughly 250–300 mg or more, in people who have gone without caffeine for days or weeks. A meta-analysis found about 109 mL of extra urine on average. Regular drinkers develop tolerance, and normal servings show little or no effect.

Does coffee count as water intake? Yes. The National Academies' reference intake is for total water from all drinks and food, 3.7 litres a day for men and 2.7 for women, and the panel concluded that caffeinated beverages can contribute as much as non-caffeinated ones. Coffee does not need to be subtracted from a daily total.

Does caffeine dehydrate you? At ordinary doses, no. A 2015 meta-analysis of 16 studies found a small increase in urine output at a median 300 mg, and no significant effect when caffeine was taken around exercise. The increase is smaller than the fluid in the drinks that normally deliver that much caffeine.

Does tea dehydrate you? No. In a trial comparing 13 drinks, a litre of hot or iced black tea containing 179 mg of caffeine produced no more urine over four hours than a litre of still water. A 2003 review estimated it would take five to eight cups at once to reach a diuretic dose.

Does decaf coffee dehydrate you? There is no reason to think so. The FDA lists decaf at about 2 to 15 mg of caffeine per 8-ounce cup, far below any dose linked to increased urine output. Decaf was not tested directly in the main hydration trial, which the authors acknowledge as a limitation.

How much coffee is dehydrating? No amount of normal-strength coffee produced net dehydration in the studies reviewed here. A diuretic effect becomes detectable at about 250–300 mg of caffeine in one sitting in people not used to it, roughly one strong to three weak 12-ounce cups on FDA figures. Intakes well above 400 mg a day have not been properly tested.

Does coffee make you pee more? It can, mostly because it is a drink: any fluid increases urine. The caffeine adds a little on top in people without tolerance, about 16% in pooled trials. In habitual drinkers on four cups a day, 24-hour urine volume was the same as on water, 2,409 versus 2,428 mL.

Does espresso dehydrate you? The studies here did not test espresso separately, so there is no direct measurement. The effect depends on the caffeine dose, not the brewing method. A drink that stays below about 250 mg of caffeine would not be expected to raise urine output measurably, although a small shot also supplies very little fluid.

Should you drink water with coffee? Only if you want to. The trial evidence does not show a fluid deficit that needs correcting, so a glass of water with coffee is a matter of taste and habit. It adds to total water intake in the ordinary way, as the coffee does.

This article summarises published research on caffeine and fluid balance for general information and is not medical advice. People who are pregnant, have kidney, heart or bladder conditions, or take diuretic medication should follow their clinician’s guidance on fluids and caffeine.