We examined the hypothesis that elevation of the body core temperature threshold for forearm skin vasodilation (TH(FVC)) with increased exercise intensity is partially caused by concomitantly increased plasma osmolality (P(osmol)). Eight young male subjects, wearing a body suit perfused with warm water to maintain the mean skin temperature at 34 +/- 1 degree C (ranges), performed 20-min cycle-ergometer exercise at 30% peak aerobic power (VO2(peak)) under isoosmotic conditions (C), and at 65% VO2(peak) under isoosmotic (H(EX)I(OS)) and hypoosmotic (H(EX)L(OS)) conditions. In H(EX)L(OS), hypoosmolality was attained by hypotonic saline infusion with DDAVP, a V2 agonist, before exercise. P(osmol) (mosmol/kg H2O) increased after the start of exercise in both H(EX) trials (P < 0.01) but not in C. The average P(osmol) at 5 and 10 min in H(EX)I(OS) was higher than in C (P < 0.01), whereas that in H(EX)L(OS) was lower than in H(EX)I(OS) (P < 0.01). The change in TH(FVC) was proportional to that in P(osmol) in every subject for three trials. The change in TH(FVC) per unit change in P(osmol) (deltaTH(FVC)/deltaP(osmol), degrees C x mosmol(-1) x kg H2O(-1)) was 0.064 +/- 0.012 when exercise intensity increased from C to H(EX)I(OS), similar to 0.086 +/- 0.020 when P(osmol) decreased from H(EX)I(OS) to H(EX)L(OS) (P > 0.1). Moreover, there were no significant differences in plasma volume, heart rate, mean arterial pressure, and plasma lactate concentration around TH(FVC) between H(EX)I(OS) and H(EX)L(OS) (P > 0.1). Thus the increase in TH(FVC) due to increased exercise intensity was at least partially explained by the concomitantly increased P(osmol).