Exposure to acute hypoxia is associated with changes in body fluid homeostasis and plasma volume (PV). This study compared a dye dilution technique using Evans' blue (PV[Evans']) with a carbon monoxide (CO) rebreathing method (PV[CO]) for measurements of PV in ten normal subjects at sea level and again 24 h after rapid passive ascent to high altitude (4,350 m). Hypobaric hypoxia decreased arterial oxygen saturation to 79 (74-83)% (mean with 95% confidence intervals). The PV(Evans') remained unchanged from 3.49 (3.30-3.68) l at sea level to 3.46 (3.24-3.68) l at high altitude. In contrast PV(CO) decreased from 3.39 (3.17-3.61) l at sea level to 3.04 (2.75-3.33) l at high altitude (P < 0.05). Compared with sea level, this resulted in an increase of the mean bias between the two methods [from 0.11 (-0.05-0.27) l at sea level to 0.43 (0.26-0.60) l at high altitude] so that the ratio between PV(Evans') and PV(CO) increased from 1.04 (0.99-1.09) at sea level to 1.15 (1.06-1.24) at high altitude (P < 0.05). In conclusion, the two methods were not interchangeable as measures of hypoxia-induced changes in PV. The mechanism responsible for the bias remains unknown, but it is suggested that the results may reflect a redistribution of albumin caused by the combined effects in hypoxia of both an increased capillary permeability to albumin and a decrease in PV. As a result, the small perivascular compartment of albumin beyond the endothelium may increase without changes in the overall albumin distribution volume.