A double-blind crossover trial was performed to assess the effect of metoprolol in 10 patients (mean age 55 years) with severe dilated cardiomyopathy. All patients clinically had idiopathic dilated cardiomyopathy; however, at coronary angiography, four had occult coronary disease. All were in New York Heart Association functional class III with a left ventricular ejection fraction less than 35% as assessed by rest radionuclide ventriculography. Studies were performed before treatment, after 4 weeks of metoprolol therapy and after 4 weeks of placebo administration. Erect bicycle sprint exercise was used to determine maximal work load. Hemodynamic variables and radionuclide left ventricular ejection fraction were recorded at rest and during graded supine bicycle exercise. Cardiac medications were unchanged throughout the trial. The mean (+/- standard error of the mean) dose of metoprolol was 130 +/- 13 mg/day. Metoprolol did not change symptoms, chest X-ray findings or exercise tolerance (baseline 700 +/- 73, placebo 690 +/- 85, metoprolol 710 +/- 81 kilopond-meters [kpm]/min). Metoprolol produced a significant decrease in heart rate at rest and during exercise (p less than 0.001). Mean blood pressure and left ventricular filling pressure did not differ significantly in the baseline, placebo and metoprolol studies. There was a slight, but significant (p less than 0.05) decrease in cardiac index with metoprolol compared with placebo and baseline studies. The small, but significant increase in left ventricular ejection fraction from baseline to the metoprolol and placebo studies (p less than 0.001) was considered a result of spontaneous improvement rather than of therapy. No significant differences were found between the patients with and without coronary disease.(ABSTRACT TRUNCATED AT 250 WORDS)