Acebutolol et Prévention Secondaire de l'Infarctus (APSI), a randomized, placebo-controlled trial, was designed to test long-term acebutolol, 200 mg twice daily, a beta blocker with mild intrinsic sympathomimetic activity, in the prevention of late death in high-risk postacute myocardial infarction (AMI) patients. APSI was planned because patients with a death rate greater than 20% have not been enrolled in significant numbers in previous trials and in such high-risk patients, it remained to be proven that beta blockers have a beneficial effect. Patients with an expected average risk of greater than 20% were to be selected based on clinical criteria. At the time of the second interim analysis, the placebo group 1-year mortality was much lower than expected (12%). The ethical board recommended to stop the trial: 309 patients had been allocated to placebo, 298 to acebutolol. The average delay between onset of symptoms and inclusion was 10.5 days. The average follow-up was 318 days after inclusion. About the same number of patients were discontinued from study treatment in both groups. All patients were included in the analysis. There were 17 deaths in the acebutolol group and 34 in the placebo group, a 48% decrease (p = 0.019). The vascular mortality decreased by 58% (p = 0.006), the highest ever observed with a beta blocker. All cardiovascular causes of death, including congestive heart failure, were less frequent in the acebutolol group. Although the objective was not achieved, APSI patients were at a higher risk than the average of the 9 previous trials with beta blockers (12% instead of 7%). In addition, the total mortality reduction did not decrease in 9 subgroups with increasing mortality risk from 2 to 23%. APSI shows that moderately severe postAMI patients can benefit from a beta-blocking treatment and a beta-blocker with mild intrinsic sympathomimetic activity can be effective.