You may have heard a recommendation that we drink half our body weight in ounces of water every day. Is this necessary ... or desirable?

The answer is: it depends — on the climate, the season, and one's activity level and diet. An active person in southern California in summer will drink a lot more water than a sedentary person in Montreal in winter. Also, a vegetarian will typically need less water than a meat eater, because meat has a lot of salt and toxins in it that have to be excreted, and it also makes the body run warmer. But the kidneys have to filter all this water, so an unnecessarily high level of water consumption doesn't come without a cost. As Michio Kushi says in Macrobiotic Dietary Recommendations, "It is best to drink only when thirsty."

The whole concept of drinking so much water is based on a misunderstanding. As the British Medical Journal and numerous other sites point out, "The advice to drink at least eight glasses of water a day can be found throughout the popular press. One origin may be a 1945 recommendation [by the U.S. National Research Council's Food and Nutrition Board] that stated, 'A suitable allowance of water for adults is 2.5 litres daily in most instances. An ordinary standard for diverse persons is 1 millilitre for each calorie of food. Most of this quantity is contained in prepared foods.' If the last, crucial sentence is ignored, the statement could be interpreted as instruction to drink eight glasses of water a day." (Footnotes omitted. For other sites, do a search for "2.5 liters water day 1945".)

Another site, by Dov Michaeli MD, PhD, observes:

The controversial 1992 bestseller Your Body’s Many Cries for Water, which calls for a minimum of eight to 10 glasses of pure water a day (not coffee, not soda), probably played a role in spreading the myth, as has the bottled-water industry, which has exploded since the 1980s.

And the scientific evidence?

In a word: none.

Exhortations to drink more water invariably include a reference to Dr. Batmanghelidj's book. Although I haven't read his book, I understand that his conclusions were based on a study of prisoners in Iran. Much of Iran has a hot climate, and one might reasonably suppose that these prisoners were often not in good health and didn't get enough to eat. More water certainly might have helped them. But to generalize to the rest of the world?

Moreover, Dr. Batman's book apparently says that if we're even a little thirsty, we're in the initial stages of dehydration, and we should immediately drink. What's the justification for this? If we're a little hungry, does that mean we're in the initial stages of starvation, and we should immediately sit down and eat? If we're a little tired, does that mean we're in the initial stages of exhaustion, and we should immediately lie down and sleep? The body is capable of operating in a range of conditions.

It's been said that thirst is not always a reliable indicator, because in some people, the hypothalamus has lost its ability to gauge our need for water, and it can take six weeks to reset. Okay, maybe people coming off a meat-eating diet need to drink a little more water for six weeks. This might be a reasonable recommendation.

Finally, a study in the June 2008 issue of the Journal of the American Society of Nephrology reviews the classic medical advice to drink eight glasses of water a day and also concludes it's a myth, as a news article about the study points out:

Drs. Dan Negoianu and Stanley Goldfarb of the University of Pennsylvania in Philadelphia reviewed the scientific literature published on the health effects of drinking lots of water to get to the roots of widely accepted urban legends about water's health benefits. ...

"Our bottom line was that there was no real good science — or much science at all — behind these claims, that they represent probably folklore," Goldfarb said.

... "When you take in a lot of water, all you do is put out more urine but not more toxins in the urine," Goldfarb said."