Blood losses are usually corrected initially by the restitution of volume with plasma expanders and subsequently by the restoration of oxygen-carrying capacity using either a blood transfusion or possibly, in the near future, oxygen-carrying plasma expanders. The present study was carried out to test the hypothesis that high-plasma viscosity hemodilution maintains perfused functional capillary density (FCD) by preserving capillary pressure. Microvascular pressure responses to extreme hemodilution with low- (LV) and high-viscosity (HV) plasma expanders and an exchange transfusion with a polymerized bovine cell-free Hb (PBH) solution were analyzed in the awake hamster window chamber model (n = 26). Systemic hematocrit was reduced from 50% to 11%. PBH produced a greater mean arterial blood pressure than the nonoxygen carriers. FCD was higher after a HV plasma expander (70 ± 15%) vs. PBH (47 ± 12%). Microvascular pressure spanning the capillary network was higher after a HV plasma expander (16–19 mmHg) compared with PBH (12–16 mmHg) and a LV plasma expander (11–14 mmHg) but lower than control (22–26 mmHg). FCD was found to be directly proportional to capillary pressure. The use of a HV plasma expander in extreme hemodilution maintained the number of perfused capillaries and tissue perfusion by comparison with a LV plasma expander due to increased mean arterial blood pressure and capillary pressure. The use of PBH increased mean arterial pressure but reduced capillary pressure due to vasoconstriction and did not maintain FCD.