In a randomized trial of the effects on in-hospital mortality of intravenous urokinase plus heparin versus heparin alone, 2,531 patients with acute myocardial infarction in 89 coronary care units were enrolled for greater than 30 months. Patients admitted within 4 hours of the onset of pain were randomized to receive either intravenous urokinase (a bolus dose of 1 million U repeated after 60 minutes) plus heparin (a bolus dose of 10,000 U followed by 1,000 IU/hour for 48 hours) or heparin alone (infused at the same rate). Complete data were obtained in 2,201 patients (1,128 taking urokinase and 1,073 taking heparin). At 16 days, overall hospital mortality was 8% in the urokinase and 8.3% in the heparin group (p = not significant). Among patients with anterior infarction, mortality was 10.3% in the urokinase and 13.9% in the heparin group (p = 0.09; relative risk = 0.73). The incidence of major bleeding (urokinase 0.44%, heparin 0.37%) as well as the overall incidence of stroke (urokinase 0.35%, heparin 0.20%) was similar in the 2 groups. The rates of major in-hospital cardiac complications (reinfarction, postinfarction angina) were also similar.