A prospective, randomised, double-blind study was performed to compare the effects of propranolol and placebo on sudden cardiac death in a high-risk group of patients who survived acute myocardial infarction. Altogether 4929 patients with definite acute myocardial infarction were screened for inclusion: 574 (11.6%) died before randomisation, and 3795 (77%) were excluded. Five hundred and sixty patients aged 35 to 70 years were stratified into two risk groups and randomly assigned treatment with propranolol 40 mg four times a day or placebo. Treatment started four to six days after the infarction. By one year there had been 11 sudden deaths in the propranolol group and 23 in the placebo group (p less than 0.038, two-tailed test analysed according to the "intention-to-treat" principle). Altogether there were 25 deaths in the propranolol group and 37 in the placebo group (P less than 0.12), with 16 and 21 non-fatal reinfarctions respectively. A quarter of the patients were withdrawn from each group. Withdrawal because of heart failure during the first two weeks of treatment was significantly more common among propranolol-treated patients than among the controls, but thereafter the withdrawal rate was the same. The significant reduction in sudden death was comparable with that after alprenolol, practolol, and timolol, which suggests that the mechanism of prevention is beta-blockade rather than any other pharmacological property of the individual drugs.