Eighty-four patients aged less than 71 years with less than 4-hour duration acute myocardial infarction (AMI) were randomized in a multicenter study to 30 U anistreplase or heparin (single injection of 6500 IU followed by 1000 IU/hr). Early reperfusion was assessed from ECG changes (50% of sum ST decrease 2 hours postdosing) and the CK release profile (CK peak less than 16 hours after onset of symptoms, CK slope greater than 10%/hr). Reperfusion rates in patients meeting at least two criteria of reperfusion were 62.5% on anistreplase versus 27.5% on heparin. On delayed angiogram (13.7 +/- 3.4 days), patency rates were 66% with anistreplase versus 47% (NS) with heparin in 76 patients. Global LVF was similar in both groups. With anistreplase, the mean lowest fibrinogen level was 0.43 +/- 0.55 g/l, plasminogen was 20 +/- 9%, and the highest F.D.P. was 1447 +/- 548 micrograms/ml. All values recovered by hour 48. In-hospital and 1-year follow-up mortality was 7.2% (three patients) with anistreplase versus 10.2% (four patients) with heparin. Bleeding occurred in 9.7% and 5.1% of the patients (NS), respectively. No intracranial hemorrhage occurred. Thus, with combined clinical criteria or reperfusion, anistreplase is twice as efficient as heparin, has a good tolerance, and is easy to use as a single injection.