In an unblinded trial of intravenous streptokinase (SK) in early acute myocardial infarction, 11 806 patients in one hundred and seventy-six coronary care units were enrolled over 17 months. Patients admitted within 12 h after the onset of symptoms and with no contraindications to SK were randomised to receive SK in addition to usual treatment and complete data were obtained in 11 712. At 21 days overall hospital mortality was 10.7% in SK recipients versus 13% in controls, an 18% reduction (p = 0.0002, relative risk 0.81). The extent of the beneficial effect appears to be a function of time from onset of pain to SK infusion (relative risks 0.74, 0.80, 0.87, and 1.19 for the 0-3, 3-6, 6-9, and 9-12 h subgroups). SK seems to be a safe drug for routine administration in acute myocardial infarction.