Study Objectives: To study the comparative effects of acute normovolemic hemodilution and nitroglycerin-induced hypotension on tissue oxygenation and blood transfusion requirement transfusion requirement. Design: Prospective, randomized study. Patients: 30 ASA physical status I and II patients scheduled for primary total hip arthroplasty. Interventions: Patients were randomized to one of three groups of 10 patients each to receive acute normovolemic hemodilutin (Group 1) or nitroglycerin-based hypotension (Group 2); Group 3 served as the control group. In Group 1, 2 U of blood was collected and replaced with an equal volume of hydroxyethyl starch (200/0.5 6%) immediately after anesthesia induction. In Group 2, nitroglycerin was infused at a rate sufficient to reduce mean arterial pressures to 60 to 65 mmHg before initiation of surgery. When hematocrit was reduced to 25%, at first autologous blood and then, if necessary, allogeneic blood was transfused to Group 1, and allogeneic blood was transfused to the other two groups, until hematocrit reached 30% for 5 days postoperatively. Measurements and Main Results: Total transfused allogeneic units of blood were determined by the fifth postoperative day. Arterial oxygen content (CaO2), venous oxygen content (CvO(2)), and oxygen extraction ratios (EO2) were calculated by standard formulas. The mean allogeneic transfusion requirement was significantly lower in Group, 1 (1.3 +/- 0.8 U) than in Group 2 (2.3 +/- 0.8 U) or Group 3 (2.7 +/- 1.1 U) (p < 0.05). In Group 1, CaO2 and CvO(2) were decreased at all times, but EO2 was significantly increased from 15 +/- 3.9% to 33.3 +/- 5.3% (p < 0.001). As for the other two groups, although CaO2 and CvO(2) were decreased, EO2 was not significantly increased. Conclusions: Acute normovolemic hemodilution is more effective than nitroglycerin-induced hypotension in reducing allogeneic blood transfusion requirement in total hip replacement surgery, without significant metabolic changes.