Some studies have demonstrated that left ventricular (LV) diastolic function is the principal determinant of impaired exercise capacity in hypertrophic cardiomyopathy (HC). In this study we sought the capability of echocardiographic indexes of diastolic function in predicting exercise capacity in patients with HC. We studied 52 patients with HC while they were not on drugs;12 of them had LV tract obstruction at rest. Diastolic function was assessed by M-mode and Doppler echocardiography by measuring: (1) left atrial fractional shortening, and the slope of posterior aortic wall displacement during early atrial emptying on M-mode left atrial tracing; and (2) Doppler-derived transmitral and pulmonary venous flow velocity indexes. Exercise capacity was assessed by maximum oxygen consumption by cardiopulmonary test during cycloergometer upright exercise. Maximum oxygen consumption correlated with the left atrial fractional shortening (r = 0.63, p <0.001), the slope of posterior aortic wall displacement during early atrial emptying (r = 0.55, p <0.001), age (r = -0.50; p <0.001), pulmonary venous diastolic anterograde velocity (r = 0.41, p <0.01), and the systolic filling fraction (r = -0.43; p <0.01). By stepwise multiple linear regression analysis, left atrial fractional shortening and the pulmonary venous systolic filling fraction were the only determinants of the maximum oxygen consumption (multiple r = 0.70; p <0.001). Exercise capacity did not correlate with Doppler-derived transmitral indexes. Thus, in patients with HC, exercise capacity was determined by passive LV diastolic function, as assessed by the left atrial M-mode and Doppler-derived pulmonary venous flow velocities.