Intravenous beta blockers given early in acute myocardial infarction have been shown to reduce chest pain, enzyme release and incidence of arrhythmias. Data published before the first report of the ISIS-1 group (International Studies of Infarct Survival) showed a 12% decrease in the probability of death using intravenous beta blockade but with large confidence limits. This study assessed and compared the effects of early intravenous atenolol and control treatment in the first week and first year after acute myocardial infarction in 16,027 patients. The principal endpoint was vascular death; most of the 15% decrease in mortality occurred in the first 24 to 36 hours with a significant difference at 1 week (313 vs 365 deaths in the atenolol and control groups, respectively). There was no rebound effect after stopping treatment after 7 days. Mortality at 1 year also showed a significant difference in favor of the atenolol group. Subgroup analysis found no significant difference in mortality by age, sex, initial heart rate, diabetes or entry electrocardiogram but data-derived analysis revealed a highly significant decrease in mortality if treatment began within 2 hours of the onset of pain. There was a significant 1% to 2% excess in inotrope use in the atenolol group in the first 36 hours, and first-degree heart block and bundle branch block emerged as relative contraindications to this treatment. The results suggest that early intravenous beta blockade in acute myocardial infarction is safe and effective and also cost effective in comparison with postdischarge oral beta blockade therapy.