To assess cardiac function and exercise tolerance in patients receiving long term TSH-suppressive therapy with levothyroxine (L-T4), we studied maximal exercise capacity with a bicycle ergometer and left ventricular function at rest and during physical exercise by radionuclide angiography. The evaluation was performed in 10 patients receiving L-T4 therapy (2.31 +/- 0.13 microgram/kg) for 5-9 yr, presenting with effort dyspnea and symptoms of adrenergic overactivity, and 10 matched control subjects. The patients were reassessed after 4 months of administration of the selective beta-adrenergic blocker bisoprolol (4.25 +/- 0.4 mg/day); L-T4 therapy remained unchanged. The results showed that at rest, left ventricular diastolic filling was impaired in the patients (P < 0.05), whereas systolic function was unaltered. During submaximal physical exercise, left ventricular ejection fraction increased in the controls from 58 +/- 2% to 65 +/- 2% (P < 0.001), whereas in the patients it fell from 63 +/- 2% to 53 +/- 2% (P < 0.01), mainly because of increased end-systolic left ventricular volume (P < 0.05). Exercise capacity was markedly reduced in the patients in terms of both peak workload (P < 0.001) and exercise duration (P < 0.001). beta-Adrenergic blockade prevented both the fall in ejection fraction and the increase in end-systolic volume during exercise, and improved exercise tolerance. In conclusion, our data show that long term TSH-suppressive therapy with L-T4 is not as harmless as believed, because it may cause marked impairment of cardiac functional reserve and physical exercise capacity. Administration of a beta-blocking drug for 4 months caused significant improvement of cardiac performance and exercise tolerance.