Responses to exchange transfusion with red blood cells (RBCs) containing methemoglobin were studied in an acute isovolemic hemodiluted hamster window chamber model to determine whether oxygen content participates in the regulation of systemic and microvascular conditions during extreme hemodilution. Two isovolemic hemodilution steps were performed with 6% dextran 70 kDa until systemic hematocrit (Hct) was reduced to 18% (Level 2). A third step hemodilution reduced the functional Hct to 75% of baseline using either a plasma expander (Dex70) or blood adjusted to 18% Hct with all RBCs containing methemoglobin (MetRBC). In vivo functional capillary density (FCD), microvascular perfusion, and oxygen distribution in microvascular networks were measured by noninvasive methods. Methylene blue was administered intravenously to reduce methemoglobin (rRBC), which increased oxygen content with no change in Hct or viscosity from MetRBC. Final blood viscosities after the entire protocol were 2.1 for Dex70 and 2.8 cP for MetRBC (baseline: 4.2 cP). MetRBC had a greater mean arterial pressure (MAP) than Dex70. FCD was substantially higher for MetRBC (82 ± 6 % of baseline) vs. Dex70 (38 ± 10 % of baseline), and reduction of methemoglobin to oxyhemoglobin did not change FCD (84 ± 5 % of baseline). PO₂ levels measured with Pd-porphyrin phosphorescence were significantly changed for Dex70 and MetRBC compared to Level 2 (Hct 18%). Reduction of methemoglobin to oxyhemoglobin partially restored PO₂s to Level 2. Wall shear rate (WSR) and wall shear stress (WSS) decreased in arterioles and venules for Dex70 and did not change for MetRBC or rRBC. Increased MAP and shear stress mediated factors could be the possible mechanisms that improved perfusion flow and FCD after exchange for MetRBC. Thus, the fall in systemic and microvascular condition during extreme hemodilution with low viscosity plasma expanders seems to be, in part, from the decrease in blood viscosity independent of the reduction in oxygen content.