In a controlled, randomised, double-blind study to see whether knowledge of left-ventricular ejection fraction (LVEF) could reduce the frequency of left-sided heart failure after acute myocardial infarction, LVEF was determined a few days before hospital discharge in a consecutive series of 60 patients. Subsequently, the patients were randomly assigned to two groups. The cardiologist responsible for their treatment was aware of the LVEF result in group I but not in group II. A month after hospital discharge there was no significant difference in the LVEF between the groups. 2 months after discharge there were no significant differences between the groups in clinical and radiological signs of left-ventricular heart failure or the use of drugs. The cardiologist's clinical estimate of the LVEF and the result of the radionuclide determination were significantly correlated. Thus, the use of LVEF did not change the clinical outcome. The need for randomised controlled studies in the evaluation of diagnostic methods is emphasised.