While long term beta-adrenergic blockade, introduced in the convalescent stage of myocardial infarction, may reduce subsequent mortality, the value of early beta-blockade in the acute phase is less certain. Therefore, the influence of beta-blockade on left ventricular performance and eventual infarct size was assessed in 61 consecutive patients with acute myocardial infarction. Metoprolol (15 mg i.v. followed by 200 mg day-1 orally) or placebo was administered in a double-blind, randomised fashion with a median delay of 5.9 hours from onset of symptoms. After 15 days of double blind therapy all patients were started on open treatment with metoprolol. All patients underwent haemodynamic monitoring for 24 hours and serial radionuclide ventriculography and thallium 201 scintigraphy. In the first hour metoprolol produced a decrease in cardiac output (1.3 l min-1; P less than 0.001) due to a reduction in heart rate (15 min-1; P less than 0.001) and a decrease in left ventricular stroke work index (10.7 g m m-2; P less than 0.001) due to a reduction in mean arterial pressure (10 mmHg; P less than 0.001). There was then a gradual attenuation in these changes. While metoprolol produced an increase in pulmonary capillary wedge pressure and in both end-diastolic and end-systolic volumes (P less than 0.05), these changes were confined to patients with a baseline pulmonary capillary wedge pressure below the median of 13 mmHg mercury. There was no significant change in stroke volume or in ejection fraction in response to metoprolol. There was no significant difference between the groups in left ventricular performance, as assessed by radionuclide ventriculography, or in scintigraphic infarct size, either at the end of the 15 days double-blind treatment or after 3 months open treatment with metoprolol. Thus, early intervention with metoprolol in acute myocardial infarction appeared to reduce myocardial oxygen consumption with no adverse haemodynamic effect. However, metoprolol failed to preserve left ventricular function, or to reduce apparent infarct size. These data suggest that the modest reduction in mortality reported in the acute phase studies of beta-blockade in myocardial infarction, is unlikely to be due to infarct reduction. It is more likely to be due to a secondary prevention or to an antiarrhythmic effect.