Sixty consecutive patients undergoing elective open-heart surgery were prospectively enrolled in a study to compare the efficacy of 3 different antifibrinolytic drugs to reduce postoperative bleeding and to reduce homologous blood requirements in combination with blood-saving techniques and restrictive indications for blood transfusion. The patients were randomized to 1 of 4 intraoperative treatment regimens: 1) control (no antifibrinolytic therapy); 2) epsilon-aminocaproic acid (10 g IV at induction of anesthesia, followed by infusion of 2 g/h for 5 hours); 3) tranexamic acid (10 mg/kg IV within 30 minutes after induction of anesthesia, followed by infusion of 1 mg/kg per hour for 10 hours); or 4) high-dose aprotinin (2 million KIU IV at induction of anesthesia and 2 million KIU added to the extracorporeal circuit, followed by infusion of 500 thousand KIU/h during surgery). Hemoconcentration and reinfusion of blood drained from the operative field and the extracorporeal circuit after operation were used in all patients. Indications for blood transfusion were hypotension, tachycardia, or both, with hemoglobin values < 8.5 g/dL; or severe anemia with hemoglobin values < 7 g/dL. Compared with the blood loss in the control group, patients receiving aprotinin and epsilon-aminocaproic acid showed significantly less postoperative blood loss at 1 hour (control, 128 +/- 94 mL; aprotinin, 54 +/- 47 mL, p = 0.01; and epsilon-aminocaproic acid, 69 +/- 35 mL, p = 0.03); this trend continued at 24 hours after operation (control, 724 +/- 280 mL; aprotinin, 344 +/- 106 mL, p < 0.0001; and epsilon-aminocaproic acid, 509 +/- 148 mL, p = 0.01). Aprotinin was significantly more efficient than epsilon-aminocaproic acid (p=0.002). Tranexamic acid did not have a statistically significant effect on blood loss. Homologous blood requirements were not significantly different among the groups; postoperative hematologic values and coagulation times were also comparable. Despite the efficacy of aprotinin and epsilon-aminocaproic acid shown in the present study, the blood requirements were not significantly different from those that are found when transfusions are restricted, autotransfusions are used, and blood from the operative field and extracorporeal circuit is concentrated and reinfused. Therefore, intraoperative antifibrinolysis may not be indicated in routine cardiac surgery when other blood-saving techniques are adopted.