IRS II (Invasive reperfusion study II) was a multicentre randomized trial comparing the efficacy of a 2-5-min 30 U anistreplase intravenous injection with a 1,500,000 U 60-min streptokinase (SK) intravenous infusion in acute myocardial infarction. 116 patients were randomized within 6 h of onset of symptoms. Early coronary patency was assessable in 107 patients by coronary angiogram performed 102 min after thrombolytic treatment (range: 30-297 min) in the anistreplase group and 93 min (range: 22-330 min) in the SK group. The early coronary patency rate was significantly higher in the anistreplase group than in the SK group: respectively, 70% (38/54) and 51% (27/53), P less than 0.05). Fifty patients had assessable coronary angiograms at 90 min and 24 h. The 24-h patency rate was 92.3% (24/26) in the anistreplase group vs 87.5% (21/24) in the SK group. No early reocclusion occurred in the anistreplase group vs 15.4% (2/13) in the SK group (NS). Fibrinogen fell to 13.2 +/- 19.8% on anistreplase vs 9.4 +/- 10.3% on SK (NS). Bleeding complications occurred in 12% (7/58) of treated patients in the anistreplase group vs 20.7% (13/58) in the SK group (NS). Two cerebrovascular accidents occurred after thrombolytic treatment with anistreplase (3.4%) vs one after SK (1.7%) (NS). Thus, anistreplase is more effective than intravenous SK and easier to administer.