BACKGROUND An oxygen-transporting hemoglobin solution should be more effective than a nonhemoglobin solution for resuscitation from hemorrhagic shock. A way to evaluate this effectiveness is to determine whether a hemoglobin solution can reverse the base deficit accumulated during hemorrhage at a faster rate than a nonhemoglobin solution. Using this criterion, we compared the resuscitative powers of autologous blood, hetastarch (Het), and diaspirin cross-linked hemoglobin (DCLHb). METHODS Fifteen sedated, spontaneously breathing sheep (37.5 +/- 10.2 kg) were bled until base deficits fell to -5 to -10 mEq/L, and plasma lactate concentrations rose to 6 to 9 mg/L. The animals were resuscitated with autologous blood (n = 5), Het (n = 5), or DCLHb (n = 5) (3.5-4.0 mL/kg every 15 minutes) until base deficits returned to prehemorrhage baseline. RESULTS Exsanguination to target base deficits required removal of an average of 41.4 +/- 5.5 mL blood/kg (estimated total blood volume, 80 mL/kg). Resuscitation required 18 +/- 3, 38 +/- 2 (different from blood), and 35 +/- 1 (different from blood) mL/kg of autologous blood, Het and DCLHb, respectively, over periods of 78 +/- 8, 163 +/- 10 (different from blood), and 129 +/- 9 minutes (different from blood and different from Het (p < or = 0.05)). Based on regression analysis, autologous blood, Het, and DCLHb corrected the base deficit at rates of, respectively, 0.074 (different from Het (p < or = 0.05)), 0.016, and 0.056 (different from Het (P < or = 0.05)) mEq/L/min. CONCLUSIONS Based on the rate of base deficit correction and the volume of solution required, autologous blood was the most effective resuscitation solution. However, DCLHb was more effective than Het. DCLHb may be an attractive alternative to blood for resuscitation from hemorrhagic shock.