Three rations, 350 ml. distilled water, 250 ml. distilled water plus 96.5 g carbohydrate, and 350 ml. distilled water plus 150 ml. sea water, were given daily for 3-day periods to six subjects receiving no other food or drink. The experiment was fully 'crossed-over' and was carried out in a constant environment. On the carbohydrate ration the water balance over the third day of exposure was about 200 ml. better than on the ration consisting of water only, and the rise in the total osmotic pressure of the body was smaller. The improvement in water balance was the result of a reduction in urine volume, which was in turn due to the effects of carbohydrate upon metabolism. These effects were (1) the sparing of body protein, (2) the prevention of ketosis and (3) a reduction of the basal metabolic rate. It is suggested that all three may have been brought about by a common mechanism. On the sea-water ration the water balance for the third day was improved by 80 to 150 ml., but the rise in the osmotic pressure of the body was greater than when distilled water alone was given. These effects were due to the retention of most of the water of the sea water, together with the salt which it contained. It is suggested that the effects of sea-water drinking on body tonicity and fluid distribution are deleterious, and that the gain of water observed on the third day would eventually have been replaced by a loss.