We subjected anesthetized mechanically ventilated rabbits (n = 6) to sequential exchanges of blood for a 6% dextran solution and compared their responses with those obtained in a previous study on progressive hypoxemia (n = 7). Right atrial PO2 (PV(O2)) (RA) and hindlimb PO2 (PV(O2))(limb), measured at the level of the iliac bifurcation, were compared with tissue PO2 (Pti(O2)) histograms obtained with an array of surface microelectrodes placed over the biceps femoris muscle. Systemic O2 consumption (V̇O2) was measured with the expired gas method. Cardiac output and systemic O2 transport (ṪO2) were calculated. Six exchanges of blood for dextran produced decreases in hemoglobin from 10.8 ± 0.4 to 2.7 ± 0.2 g/dl (P < 0.001). Critical ṪO2 (ṪO2(crit)), defined as the level of ṪO2 associated with initial decreases in control V̇O2, was similar for anemia and hypoxemia (40.5 ± 5.6 and 40.1 ± 5.3 ml · min-1 · kg-1, respectively). At any given ṪO2 other than control ṪO2, the levels of (PV(O2))(RA) and (PV(O2))(limb) were greater in anemia than in hypoxemia (P < 0.01), but the mean and the distribution of the Pti(O2) histograms were similar in both conditions. Mean Pti(O2) was significantly less than (PV(O2))(RA) or (PV(O2))(limb), except for those values obtained during the control period. These results confirm our previous finding that PV(O2) is not an accurate index of Pti(O2) under conditions of tissue hypoxia. Furthermore, similar Pti(O2) levels during anemia and hypoxemia suggest that V̇O2 is limited by decreases in O2 diffusion from the capillaries to the cells.