The effects of normothermic systemic perfusion (35 degrees to 37 degrees C; n = 73) were compared with those of moderately hypothermic systemic perfusion (25 degrees to 29 degrees C; n = 73) with respect to blood loss, transfusion requirements, and platelet levels in 146 patients undergoing isolated, primary coronary artery bypass grafting. In addition, most patients were given an antifibrinolytic medication during operation as follows: tranexamic acid (10 gm intravenously; n = 63), epsilon-aminocaproic acid (15 gm intravenously; n = 63), or no drug as a control. (n = 20). Normothermic patients tended to bleed less at 24 hours (warm, 864 +/- 42 ml and cold, 918 +/- 68 ml), but these differences were not statistically significant. Patients receiving either tranexamic acid or epsilon-aminocaproic acid, regardless of perfusion temperature, bled less after 6, 12, and 24 hours than did cold control patients (p less than 0.05). Warm control patients also bled less than did cold control patients after 6 or 12 hours (p less than 0.05), and neither drug further reduced blood loss in these patients. Circulating platelet levels were better preserved in patients receiving either tranexamic acid or epsilon-aminocaproic acid and in patients with warm perfusion and no drug than in cold control patients. Normothermic systemic perfusion, tranexamic acid, and epsilon-aminocaproic acid each reduced postoperative blood loss and preserved platelets.