All patients with a presumptive diagnosis of myocardial infarction, who were seen within six hours of the onset of symptoms and had no reason for exclusion, were considered for entry into a trial to compare the effects of metoprolol and placebo on creatine kinase MB isoenzyme release. The trial was randomised and double blind. The median time from onset of symptoms to receiving trial drug was just under two hours. Two hundred and four patients (of whom 120 had myocardial infarction) received metoprolol and 187 (of whom 105 had myocardial infarction) received placebo. Infarct size was estimated semiquantitatively using cumulative release of the cardiospecific isoenzyme, creatine kinase MB. Mean creatine kinase MB isoenzyme was less in patients who received metoprolol, but the reduction did not achieve statistical significance. Clinical problems related to early intravenous metoprolol were uncommon.