To determine how the sodium content of ingested fluids affects drinking and the restoration of the body fluid compartments after dehydration, we studied six subjects during 4 h of recovery from 90-100 min of a heat [36°C, <30% relative humidity (rh)] and exercise (40% maximal aerobic power) exposure, which caused body weight to decrease by 2.3%. During the 1st h, subjects rested seated without any fluids in a thermoneutral environment (28°C, <30% rh) to allow the body fluid compartments to stabilize. Over the next 3 h, subjects rehydrated ad libitum using tap water and capsules containing either placebo (H2O-R) or 0.45 g NaCl (Na-R) per 100 ml water. During the 3-h rehydration period, subjects restored 68% of the lost water during H2O-R, whereas they restored 82% during Na-R (P < 0.05). Urine volume was greater in H2O-R than in Na-R; thus only 51% of the lost water was retained during H2O-R, whereas 71% was retained during Na-R (P < 0.05). Plasma osmolality was elevated throughout the rehydration period in Na-R, whereas it returned to the control level by 30 min in H2O-R (P < 0.05). Changes in free water clearance followed changes in plasma osmolality. The restoration of plasma volume diring Na-R was 174% of that lost. During H2O-R it was 78%, which seemed to be sufficient to diminish volume-dependent dipsogenic stimulation. These results suggest that the poorer rehydration when drinking water is caused by both removal of the osmotic drive for drinking and a rise in free water clearance, primarily due to the loss of electrolytes during dehydration. In addition, the higher degree of recovery in plasma volume than in total body water during H2O-R and Na-R delayed rehydration by removing the volume-dependent dipsogenic stimulation.