Background Prophylactic administration of the antifibrinolytic drug tranexamic acid decreases bleeding and transfusions after cardiac operations. However, the best dose of tranexamic acid for this purpose remains unknown. This study explored the dose-response relationship of tranexamic acid for hemostatic efficacy after cardiac operation. Methods In prospective, randomized, double-blinded fashion, 148 patients undergoing cardiac operation with extracorporeal circulation were divided into six groups: a placebo group and five groups receiving tranexamic acid in loading doses before incision (range 2.5 to 40 mg.kg-1) and one-tenth the loading dose hourly for 12 h. The mass of blood collected by chest tubes over 12 h represented blood loss. Allogeneic transfusions within 12 h and within 5 d of surgery were tallied. Results The six groups presented similar demographics. Patients receiving placebo had increased postoperative D-dimer concentration compared to groups receiving tranexamic acid. Patients receiving at least 10 mg.kg-1 tranexamic acid followed by 1 mg.kg-1.h-1 bled significantly less (365, 344, and 369 g.12 h-1, respectively, for those three groups) compared with patients who received placebo (552 g, P < 0.05). Tranexamic dose did not affect transfusions. Only initial hematocrit affected whether a patient received an allogeneic transfusion within 5 days of operation (odds ratio 2.08 for each 3% absolute decrease in hematocrit). Conclusions Prophylactic tranexamic acid, 10 mg.kg-1 followed by 1 mg.kg-1.h-1, decreases bleeding after extracorporeal circulation. Larger doses do not provide additional hemostatic benefit.