Postoperative hemorrhage in patients undergoing open-heart surgery is a major cause of morbidity and mortality. Monitoring of coagulation in these patients has routinely involved the activated clotting time. Thromboelastography is currently used as a monitor of coagulation during liver transplantation. The thromboelastogram, by providing information on the interaction of all the coagulation precursors, gives more clinically useful information on coagulation than that available from the coagulation profile or the activated clotting time alone. This study was done to assess the usefulness of thromboelastography in open-heart surgery. Thirty-eight patients (29 undergoing coronary artery bypass grafting and 9 undergoing valve replacement) were studied with activated clotting time, thromboelastography, and coagulation profiles during three periods: before bypass, during bypass, and after protamine administration. Thromboelastography was a significantly better predictor (87% accuracy) of postoperative hemorrhage and need for reoperation than was the activated clotting time (30%) or coagulation profile (51%). Thromboelastography is easy to use and provides diagnostic data within 30 minutes of blood sampling.