The effect of oral propranolol on left ventricular ejection fraction, left ventricular volumes, cardiac output, and segmental wall motion was assessed with multigated blood pool imaging both at rest and during supine exercise in 15 patients with angina pectoris. Propranolol had no effect on resting left ventricular ejection fractions. Before propranolol, they did not change during exercise, whereas after propranolol the ejection fractions increased slightly. Exercise left ventricular ejection fractions increased with propranolol in three patients with resting left ventricular ejection fractions of less than 40 per cent. More specifically, left ventricular end-diastolic volume index, end-systolic volume index, stroke volume index, and cardiac index were not altered significantly at rest or during exercise by propranolol. Exercise left ventricular ejection fractions were increased in five and unchanged in eight patients by propranolol. Those patients with increases in left ventricular ejection fractions had a greater change in left ventricular end-diastolic volume indices and a greater change in left ventricular end-systolic volume indices during exercise while on propranolol. Left ventricular segmental wall motion was not altered significantly during exercise by propranolol. We conclude that: (1) Left ventricular functional responses to propranolol during exercise are heterogeneous and not easily predicted; (2) propranolol causes no consistent deterioration in exercise left ventricular ejection fraction even in patients with resting ventricular ejection fractions less than 40 per cent; (3) increased exercise left ventricular ejection fraction with propranolol is contributed to by significant increases in end-diastolic volume during exercise; and (4) gated blood pool imaging is a useful method for characterising rest and exercise left ventricular ejection fractions and left ventricular volumes during propranolol therapy.