A double-blind controlled trial was performed to evaluate the dose-effect relationship of APSAC on coronary reperfusion in patients with acute myocardial infarction. A total of 103 patients were recruited in 14 centres and randomly allocated to receive 3.75, 7.5, 15.0 or 30.0 U of anisoylated plasminogen streptokinase activator complex (APSAC), a thrombolytic. Angiograms were taken at baseline to assess eligibility and at six times after administration at 15-min intervals (up to 90 min). A control angiogram was performed at 24 h for patients with a patent artery at 90 min. Angiograms were read centrally by independent readers. There was a non-significant difference between the four groups for the main outcome, reperfusion of an initially occluded artery: 5/20 (25%; 3.75 U); 13/22 (59%; 7.5 U); 11/22 (50%; 15.0 U); 14/25 (56%; 30.0 U); p = 0.11. There was no apparent difference between group 7.5 U, 15.0 U and 30.0 U (mean recanalisation rate: 55%), but the reperfusion rate in the 3.75 U group was significantly lower (p = 0.02). No apparent differences were observed for the delay of recanalisation between the four groups. The mean values of minimum fibrinogen concentration were different in the four groups, with a clear dose-effect relationship (p = 0.0004). The side effects were not found to be dose-dependent.