BACKGROUND In acute myocardial infarction patients who do not reperfuse their infarct arteries shortly after thrombolytic treatment have a high morbidity and mortality. Management of this high risk group remains problematic, especially in centres without access to interventional cardiology. Additional thrombolytic treatment may result in reperfusion and improved left ventricular function. METHODS Failure of reperfusion was assessed non-invasively as less than 25% reduction of ST elevation in the electrocardiographic lead with maximum ST shift on a pretreatment electrocardiogram. 37 patients with acute myocardial infarction who showed electrocardiographic evidence of failed reperfusion 30 minutes after 1.5 MU streptokinase over 60 minutes were randomly allocated to receive either alteplase (tissue type plasminogen activator (rt-PA) 100 mg over three hours) (19 patients) or placebo (18 patients). 43 patients with electrocardiographic evidence of reperfusion after streptokinase acted as controls. Outcome was assessed from the Selvester Q wave score of a predischarge electrocardiogram and a nuclear gated scan for left ventricular ejection fraction 4-6 weeks after discharge. RESULTS Among patients in whom ST segment elevation was not reduced after streptokinase, alteplase treatment resulted in a significantly smaller electrocardiographic infarct size (14% (8%) v 20% (9%), P = 0.03) and improved left ventricular ejection fraction (44 (10%) v 34% (16%), P = 0.04) compared with placebo. This benefit was confined to patients who failed fibrinogenolysis after streptokinase (fibrinogen > 1 g/l). In patients in whom ST segment elevation was reduced after streptokinase, infarct size and left ventricular ejection fraction were not significantly different from those in patients treated with additional alteplase. CONCLUSIONS Patients without electrocardiographic evidence of reperfusion after streptokinase may benefit from further thrombolysis with alteplase.