Eighteen patients with pure aortic stenosis without coronary artery disease underwent equilibrium radionuclide angiography to evaluate the adaptation of their left ventricular function to exercise. The left ventricular ejection fraction, peak left ventricular ejection, and fillings, and their timing were calculated from time-activity curves and their first derivatives at rest and at the maximum of exercise. There were no clinical complications. The ST segment and T wave changes of 14 patients were accentuated and 3 patients developed anginal pain. The ejection fraction was normal at rest and did not change significantly during exercise. The peak ejection did not vary but peak left ventricular filling was prolonged by exercise. There was a correlation between peak ventricular ejection at rest and the aortic valve surface area at catheterisation. This isotopic parameter was inversely correlated with LVEDP. There was a close correlation between age and peak ventricular filling on exercise. The variation between resting and exercise values of this parameter was inversely correlated with age. This study shows that exercise stress testing can be undertaken without risk in patients with aortic stenosis. The results of radionuclide angiography show that peak left ventricular ejection is a valuable parameter. The interpretation of the diastolic parameters is however more difficult because they are age-related.