Pump failure, ranging from ventricular dysfunction to acute cardiogenic shock, is now the leading cause of cardiac death. Efforts at temporary mechanical or pharmacologic support of the heart have been largely unsuccessful so that attention is now directed toward prevention of ventricular failure and limitation of myocardial infarct size or even outright prevention of infarction itself. In particular, attention has been refocused on earlier reperfusion efforts with streptokinase. The effect of thrombolysis in acute myocardial infarction on enzymatic infarct size, left ventricular function and early mortality was studied in subsets of patients in a randomized trial (Netherlands Interuniversity Cardiology Institute). Early thrombolytic therapy with intracoronary streptokinase (152 patients) or with intracoronary streptokinase preceded by intravenous streptokinase (117 patients) was compared with conventional treatment (264 patients). All 533 patients were admitted to the coronary care unit within 4 hours after onset of symptoms indicative of acute myocardial infarction. Of the patients eligible for this detailed analysis, 245 were allocated to thrombolytic therapy and 243 to conventional treatment. Early angiography was preformed in 212 of the 245 patients allocated to thrombolytic therapy. Patency of the infarct-related artery was achieved in 181 patients (85%). Enzymatic infarct size, measured from cumulative alpha-hydroxybutyrate dehydrogenase release, was smaller in patients allocated to thrombolytic therapy (median 760 versus 1,179 U/liter in control subjects, p = 0.0001). Left ventricular ejection fraction measured by radionuclide angiography before discharge was higher after thrombolytic therapy (median 50% versus 43% in control subjects, p = 0.0001). Twelve month mortality was lower in patients allocated to thrombolytic therapy (8% versus 16% in the control group, p less than 0.01). In multivariate regression analysis infarct size limitation, improvement of left ventricular ejection fraction and 3 month mortality were predicted by sigma ST, time from onset of symptoms to admission and Killip class at admission. Thrombolysis was most useful in patients admitted within 2 hours after onset of symptoms and in patients with a sigma ST segment of 1.2 mV or more. On the other hand, no beneficial effects of streptokinase on enzymatic infarct size, left ventricular function or mortality were observed in the subset of patients with sigma ST less than 1.2 mV, admitted 2 to 4 hours after onset of symptoms.