In a randomised, double-blind study 1258 patients were allocated to receive either anistreplase (anisoylated plasminogen streptokinase activator complex [APSAC], 'Eminase') or placebo within 6 h of onset of suspected acute myocardial infarction. At 30 days, 40 (6%) of 624 patients on anistreplase had died, compared with 77 (12%) of 634 patients on placebo (odds reduction 50.5%). On long-term follow-up a survival benefit was still observed: at 12 months, 69 (11%) patients treated with anistreplase had died, compared with 113 (18%) patients given placebo (odds reduction 43%; p = 0.0007, 95% confidence interval 21-59%). This effect on mortality was not related to time between onset of symptoms and treatment or to any patient characteristic. Site of infarction and age were the most important influences on 1-year survival in both treatment groups; tachycardia (over 100 beats/min) on admission and previously diagnosed ischaemic heart disease were also associated with increased risk. Major complications of acute myocardial infarction were less frequent in patients treated with anistreplase than in controls. As for other thrombolytic agents, haemorrhage was more common, but usually minor. These findings indicate that anistreplase is an effective and acceptably safe thrombolytic with long-term survival benefits for patients with acute myocardial infarction.