OBJECTIVE: Ecallantide is a recombinant peptide in the same class as aprotinin that inhibits plasma kallikrein, a major component of the contact coagulation and inflammatory cascades. Therefore, ecallantide was expected to reduce blood loss associated with cardiac surgery requiring cardiopulmonary bypass. METHODS: This prospective multinational, randomized, double-blind trial enrolled patients undergoing cardiac surgery using cardiopulmonary bypass for procedures associated with a high risk of bleeding. Patients were randomly assigned to ecallantide (n = 109) or tranexamic acid (high dose, n = 24; low dose, n = 85). Efficacy was assessed from the volume of packed red blood cells administered within the first 12 hours after surgery. RESULTS: The study was terminated early after the independent data safety and monitoring board observed a statistically significantly higher 30-day mortality in the ecallantide group (12%) than in the tranexamic acid groups (4%, P = .041). Patients receiving ecallantide received more packed red blood cells within 12 hours of surgery than tranexamic acid-treated patients: median = 900 mL (95% confidence interval, 600-1070) versus 300 mL (95% confidence interval, 0-523) (P < .001). Similar differences were seen at 24 hours and at discharge. Patients treated with the higher tranexamic acid dose received less packed red blood cells, 0 mL (95% confidence interval, 280-600), than the group treated with the lower dose, 400 mL (95% confidence interval, 0-400) (P = .008). No deaths occurred in the higher dose tranexamic acid group. CONCLUSIONS: Ecallantide was less effective at reducing perioperative blood loss than tranexamic acid. High-dose tranexamic acid was more effective than the low dose in reducing blood loss.