We investigated the effect of increasing hemoglobin- (Hb) O2 affinity on muscle maximal O2 uptake (Vo2max) while muscle blood flow, [Hb], HbO2 saturation, and thus O2 delivery (muscle blood flow x arterial O2 content) to the working muscle were kept unchanged from control. Vo2max was measured in isolated in situ canine gastrocnemius working maximally (isometric tetanic contractions). The muscles were pump perfused, in alternating order, with either normal blood [O2 half-saturation pressure of hemoglobin (P50) = 32.1 +/- 0.5 (SE) Torr] or blood from dogs that had been fed sodium cyanate (150 mg.kg-1.day-1) for 3-4 wk (P50 = 23.2 +/- 0.9). In both conditions (n = 8) arterial Po2 was set at approximately 200 Torr to fully saturate arterial blood, which thereby produced the same arterial O2 contents, and muscle blood flow was set at 106 ml.100 g-1.min-1, so that O2 delivery in both conditions was the same. Vo2max was 11.8 +/- 1.0 ml.min-1.100 g-1 when perfused with the normal blood (control) and was reduced by 17% to 9.8 +/- 0.7 ml.min-1.100 g-1 when perfused with the low-P50 blood (P < 0.01). Mean muscle effluent venous Po2 was also significantly less (26 +/- 3 vs. 30 +/- 2 Torr; P < 0.01) in the low-P50 condition, as was an estimate of the capillary driving pressure for O2 diffusion, the mean capillary Po2 (45 +/- 3 vs. 51 +/- 2 Torr). However, the estimated muscle O2 diffusing capacity was not different between conditions. The proportional reductions in Vo2max and calculated mean capillary Po2 are consistent with the hypothesis that O2 diffusion limitation in the peripheral tissues can be one important determinant of Vo2max. In addition, these results provide evidence that incomplete O2 extraction by the maximally working muscle is not only a result of perfusional O2 shunts or functional heterogeneity, because at constant O2 delivery, neither of these factors should affect O2 extraction as P50 is reduced.