We investigated whether the reduction in calculated muscle diffusion capacity for O2 (DmO2) previously shown to occur with lowered hemoglobin concentration ([Hb]) perfusion of maximally working muscle is related to changes in the blood flow distribution. If blood flow distribution is altered during low [Hb] conditions, the reduction in the calculated DmO2 may in fact be due to increasing heterogeneity and not to some other hemoglobin-related factor. Color-stained (15-microns- diam) microspheres were injected into the artery supplying maximally working isolated in situ dog gastrocnemius muscle (n = 6) while it was being perfused (flow controlled by pump perfusion) with whole blood at three different levels of [Hb] (14.1 +/- 0.5, 8.9 +/- 0.4, and 5.7 +/- 0.4 (SE) g/100 ml] in a blocked-order design. Muscle blood flow and arterial PO2 were not changed as [Hb] was altered. Maximal O2 uptake (11.8 +/- 1.3, 8.2 +/- 0.8, and 6.0 +/- 0.9 ml.100 g-1 min-1 for those [Hb] values, respectively) and the associated estimate of DmO2 (0.25 +/- 0.03, 0.18 +/- 0.03, and 0.15 +/- 0.03 ml.100 g-1.min-1.Torr-1) declined significantly (P < 0.05) with [Hb]. However, the dispersion of the blood flow distribution did not change significantly and, if anything, indicated less heterogeneity at lower [Hb] (coefficient of variation - 0.52 +/- 0.06, 0.46 +/- 0.05, and 0.43 +/- 0.03). These results suggest that in maximally working canine muscle in situ, when O2 delivery is reduced by lowering [Hb] (at constant blood flow), changes in blood flow distribution play no significant role in the reduction of maximal O2 uptake and calculated DmO2. The apparent increase in the resistance to O2 diffusion (i.e., reduction in the DmO2) during anemia may therefore be a result of increased red blood cell spacing in the capillary, slow chemical off-loading kinetics of O2 from Hb, or some other effect that remains to be determined.