Exercise capacity in patients with dilated cardiomyopathy, measured by peak oxygen consumption (VO₂) during exercise, has virtually no relation to resting left ventricular (LV) function. We hypothesized that exercise-induced inotropic reserve may explain some of the variation between peak VO₂ and resting LV function. Treadmill stress echocardiography was performed simultaneously with peak VO₂ measurements in 35 patients with dilated cardiomyopathy. Resting and immediate postexercise echocardiographic images were scored for change in segmental contractility using the American Society of Echocardiography 16-segment system. Segment scores were summed and divided by 16 to determine the wall motion index. Right ventricular (RV) function was quantified on a 4-point scale. Patients had a mean age of 52 ± 12 years (8 women) and a mean ejection fraction of 30 ± 10 (25 nonischemic patients). Average peak VO₂ was 17.0 ± 6 ml/kg/min. Patients were divided into 2 groups by peak VO₂: a high VO₂ group, >17 ml/kg/min (17 patients) and a low VO₂ group, ≤17 ml/kg/min (18 patients). LV ejection fraction was similar between the high and low VO₂ groups (31 ± 9% vs 28 ± 11%, p = NS) as were etiology of heart failure, medications used, and LV volume. In the high VO₂ group, wall motion index improved from 2.28 ± 0.20 to 2.12 ± 0.31 during exercise (p = 0.009). There was no improvement in the low VO₂ group. Resting RV function was significantly better in the high VO₂ group (1.4 ± 0.8 vs 0.6 ± 0.6 p = 0.004). Therefore, in patients with dilated cardiomyopathy and similar resting LV function, the presence of demonstrable LV inotropic reserve and preserved RV function partially account for variation in exercise performance.