To test the hypothesis that acute hypoxia does not modify the relationship between plasma vasopressin concentration ([AVP]_p) and plasma osmolality (P_osmol) during exercise and that the increase in [AVP]_p during exercise is due mainly to the exercise intensity-dependent increase in P_osmol, we examined [AVP]_p during a graded exercise in a hypoxic condition (13% O₂, N₂ balance) in seven healthy male subjects. A graded exercise in a normoxic condition on a separate day served as the control. Hypoxia reduced peak aerobic power (O_2 peak) by 32.4 ± 2.7%. Blood samples obtained during rest and at around 25, 45, 65, 80, and 100% of O_2 peak of each of the respective conditions were used for analyses of intravascular water and electrolyte balance. The pattern of the changes in fluid and electrolyte balance in response to percent O_2 peak was similar between the two conditions. Plasma volume decreased linearly as percent O_2 peak increased while P_osmol increased in a curvilinear fashion with a steep increase occurring at above ~66% O_2 peak. Above this relative exercise intensity, plasma sodium, potassium, and lactate concentrations also increased, whereas plasma bicarbonate concentration decreased. Thus transvascular fluid movement at above ~66% O_2 peak was due to the net efflux of hypotonic fluid out of the vascular space in both conditions. The relationship between [AVP]_p and P_osmol during exercise in response to relative exercise intensity was similar between the two conditions. The results indicate that acute mild hypoxia itself has no direct effect on vasopressin release, and it does not modify the relationship between [AVP]_p and P_osmol during exercise. The results also support the hypothesis that exercise-induced vasopressin release is primarily stimulated by increased P_osmol produced by hypotonic fluid movement out of the vascular space in a relative exercise intensity-dependent manner.