Large randomised trials have demonstrated that fibrinolytic therapy can reduce mortality in patients with suspected acute myocardial infarction (AMI). The indications for, and contraindications to, this treatment in some categories of patient are disputed, examples being late presentation, elderly patients, and those in cardiogenic shock. This overview aims to help resolve some of the remaining uncertainties. From all trials of fibrinolytic therapy versus control that randomised more than 1000 patients with suspected AMI, information was sought and checked on deaths during the first 5 weeks and on major adverse events occurring during hospitalisation. The nine trials included 58,600 patients, among whom 6177 (10.5%) deaths, 564 (1.0%) strokes, and 436 (0.7%) major non-cerebral bleeds were reported. Fibrinolytic therapy was associated with an excess of deaths during days 0-1 (especially among patients presenting more than 12 h after symptom onset, and in the elderly) but this was outweighed by a much larger benefit during days 2-35. This "early hazard" should not obscure the very clear overall survival advantage that is produced by fibrinolytic therapy. Benefit was observed among patients presenting with ST elevation or bundle-branch block (BBB)--irrespective of age, sex, blood pressure, heart rate, or previous history of myocardial infarction or diabetes--and was greater the earlier treatment began. Among the 45,000 patients presenting with ST elevation or BBB the relation between benefit and delay from symptom onset indicated highly significant absolute mortality reductions of about 30 per 1000 for those presenting within 0-6 h and of about 20 per 1000 for those presenting 7-12 h from onset, and a statistically uncertain benefit of about 10 per 1000 for those presenting at 13-18 h (with more randomised evidence needed in this latter group to assess reliably the net effects of treatment). Fibrinolytic therapy was associated with about 4 extra strokes per 1000 during days 0-1: of these, 2 were associated with early death and so were already accounted for in the overall mortality reduction, 1 was moderately or severely disabling, and 1 was not. This overview indicates that fibrinolytic therapy is beneficial in a much wider range of patients than is currently given such treatment routinely.