The aim of this study was to compare the effects of intraoperative autotransfusion and tranexamic acid on postoperative bleeding and the need for allogeneic transfusion. In a prospective randomized study, 200 patients undergoing coronary artery bypass were divided into two groups: 100 patients received 1-2 units of autologous blood after termination of cardiopulmonary bypass; and 100 patients were given tranexamic acid 15 mg x kg(-1) before injection of heparin and again before injection of protamine. Postoperative bleeding was significantly lower in the tranexamic acid group (600 mL) than the autotransfusion group (1,100 mL). The percentage of patients transfused in the autotransfusion and tranexamic acid groups was 70% and 65%, respectively. Patients in the autotransfusion group received significantly more whole blood (2.82 vs 1.93 units). Intensive care and hospital stays were shorter in the tranexamic acid group. There was no hospital mortality and no difference in thrombotic complications between groups. Tranexamic acid was more effective than autotransfusion in reducing postoperative blood loss and allogeneic transfusions after coronary bypass.