AIMS Valve compliance might determine the onset of symptoms better than resting measures of aortic stenosis. This study compared valve compliance measured by dobutamine stress echocardiography with resting haemodynamic variables against the end-point of symptoms at low workload during exercise testing. METHODS AND RESULTS Echocardiography was performed at rest and during each stage of a dobutamine stress test in 65 asymptomatic patients with moderate or severe aortic stenosis. Each patient also completed a modified Bruce treadmill exercise test. During dobutamine stress, peak transaortic velocity increased by 1.0 (0.4) m/s and effective orifice area by 0.25 (0.22) cm(2). Valve compliance was 0.23 (0.10) cm(2)/100ml.s(-1), and was independent of baseline effective orifice area. In the 19 patients limited by symptoms on exercise testing, valve compliance was significantly lower (0.19 (0.09) cm(2)/100ml.s(-1)) than in those who remained asymptomatic (0.25 (0.10) cm(2)/100ml.s(-1), p=0.03). Effective orifice area at peak stress was also lower (1.0 (0.3) vs 1.2 (0.4) cm(2), p=0.03), but there were no significant differences in resting measures of effective orifice area, transaortic velocity, or mean pressure drop. CONCLUSIONS Effective orifice area is flow-dependent in patients with moderate and severe aortic stenosis with preserved left ventricular function. Exertional symptoms are better predicted by compliance than resting effective orifice area, mean pressure drop or peak transaortic velocity.