BACKGROUND: Prophylactic administration of tranexamic acid (TA) reduces bleeding and transfusion requirement after open heart operations. This study was performed to determine the relationship between inhibition of fibrinolysis and TA blood concentration. METHOD: In phase I, recombinant tissue plasminogen activator[r-tPA (0, 50, 100, 150 ng/ml)] was added to the blood of volunteer and induced fibrinolysis. In phase II, 4 thromboelastography (TEG) models of severe fibrinolysis in which TA was added to achieve blood levels (0, 0.72, 1.44, 2.88 mg/ml) were compared to determine the lowest effective dose. In phase III, the lowest dose (0.72 mg/ml) was mixed with the blood and evaluated on TEG in open heart operation. In phase IV, a placebo group and study group receiving TA in an loading dose of 5 mg/kg before bypass following infusion of 2 mg/kg/hour. Used analysis is Mann Whitney U test and Wilcoxon rank signed test. RESULT: In phase I, fibrinolytic inhibition at A30/MA (r=0.752) and A60/MA (r=0.735) were linearly correlated with the blood r-tPA concentration. In phase II, severe fibrinolysis (r-tPA 100 ng/ml) was reversed completely at all doses of TA. In phase III, the fibrinolysis index at 10 min. after starting bypass, aorta declamping, and 1 hour after operation were improved when the patient's blood was treated with TA (0.72 mg/ml). In phase IV, blood treated with TA showed less fibrinolysis and better TEG results than the placebo group. CONCLUSION: A small dose of TA (5 mg/kg), which was determined by an in vitro model of fibrinolysis on TEG, was effective in preventing changes in fibrinolytic index during cardiopulmonary bypass in open heart surgery.