The effect of metoprolol on the development of an acute myocardial infarction (AMI) during days 0 to 3 and on late first and recurrent infarctions during days 4 to 15 has been investigated. Signs on electrocardiogram (ECG) were well balanced between the treatment groups at entry; 70% of patients had signs of suspected AMI and 19% of patients had normal ECGs. The remaining patients had abnormal ECGs but actual infarction could not be localized. The localization of suspected AMI was equivalently distributed in the 2 groups before randomization. Metoprolol altered the distribution of patients diagnosed during days 0 to 3 as having definite, possible or no AMI (p less than 0.02). In the placebo group, there were more patients with definite AMI (72.5% vs 70.5%) and less with possible AMI (5.6% vs 7.4) than in the metoprolol group. A larger proportion of patients developed a Q-wave infarction during days 0 to 3 in the placebo group (53.9%) compared with the metoprolol group (50.9%, p = 0.024). No difference in the effect of metoprolol regarding localization of the early AMI was observed. Late first myocardial infarction development (days 4 to 15) was observed in 20 patients (0.7%) in each group. Recurrent myocardial infarction tended to develop more frequently during days 4 to 15 in the placebo group compared with the metoprolol group (3.9% vs 3.0%, p = 0.08).