The perioperative consequences of preoperative aspirin administration were assessed in a large prospective cooperative study of 772 patients undergoing coronary artery bypass grafting. The 772 patients were randomized to receive either aspirin (325 mg once a day), aspirin (325 mg three times a day), aspirin plus dipyridamole (325 and 75 mg together three times a day) (aspirin group), sulfinpyrazone (267 mg three times a day) or placebo (nonaspirin group). The therapy, except in the aspirin group, was started 48 h before the operation. In all aspirin subgroups, one 325 mg aspirin dose was given 12 h before surgery and maintained thereafter according to the assigned regimen. Patients in the aspirin group had significantly more postoperative bleeding and received more packed blood cells and blood products than did patients in the nonaspirin group. Although total operative duration and cardiopulmonary bypass duration were not different, the interval between completion of cardiopulmonary bypass and wound closure was significantly longer (p = 0.035) in the aspirin group. Thirty-one (6.6%) of 471 patients in the aspirin group and 5 (1.7%) of 301 patients in the nonaspirin group also required reoperation for control of postoperative bleeding (p = 0.002). The site of bleeding found at reoperation was not different between the two groups. There was no difference in operative mortality rates, incidence of other bleeding complications or occurrence of other post-operative complications between the two groups. Thus, antiplatelet regimens involving preoperative initiation of aspirin therapy, which has been shown to improve graft patency, increase the risk of abnormal postoperative bleeding and the need for reoperation.(ABSTRACT TRUNCATED AT 250 WORDS)