In the study presented here, we examined the affects of a close to complete replacement of sweat water and Na+ losses on fluid shifts during exercise. Six cyclists performed three 4-h rides at 55% of their peak oxygen uptake in a 20 degrees C environment while consuming 3.85 l of an 8% carbohydrate solution containing 5, 50 or 100 mEq.l-1 of Na+. Increases in Na+ intake reduced renal free water clearance from around 40 ml.h-1 to -8 and -121 ml.h-1 and led to a decrease in urine volume from approximately equal to 1.0 to 0.5 l (P < 0.05). In contrast, the 3.5-3.9 l fluid and 150-190 mEq Na+ losses in sweat were similar in each trial, as were the approximately equal to 80 mEq K+ losses in sweat and urine and the 282-288 mosmol.kg-1 plasma osmolalities. During the low-Na+ trial, plasma osmolality was maintained by a approximately equal to 1.3 l contraction of extracellular fluid (ECF) with the loss of approximately equal to 200 mEq Na+. However, in the other trials, approximately equal to 1.3 l of water was lost from the intracellular fluid. During the medium-Na+ trial, a loss of only approximately equal to 40 mEq Na+ maintained ECF volume, and during the high-Na+ trial, a gain of approximately equal to 160 mEq Na+ expanded the ECF by approximately equal to 0.8 l. However, corresponding changes in plasma volumes from -0.20 to 0.15 l had no effect on cardiovascular drift or thermoregulation. These data suggest that during prolonged exercise of moderate intensity under mild environmental conditions when sweat rates are approximately equal to 0.9 l.h-1, complete Na+ replacement maintains plasma volume and reduces dehydration, but when fluid intake matches sweat rate, has little effect on plasma osmolality.