Animal studies have shown that the power output of the left ventricle is related to afterload by a bell-shaped curve. Furthermore, the normal ventricle operates at the maximum power point, whereas the diseased ventricle operates off the maximum power point because of increased afterload. We studied this relation in eight patients with dilated cardiomyopathy. A high-fidelity catheter with micromanometer pressure and electromagnetic velocity transducers was used to measure blood pressure and flow velocity in the ascending aorta. The latter was converted into volumetric flow by calibrating with the simultaneously determined thermodilution cardiac output. Ventricular power was calculated by dividing the integral of the aortic blood pressure-flow product by the duration of the cardiac cycle. Intravenous nitroprusside was used to alter afterload and the power-afterload relation was obtained by plotting power against mean aortic blood pressure. In all patients, as blood pressure was lowered initially, the power output of the left ventricle increased. As the dose of nitroprusside was increased further, the total power either plateaued or actually decreased. The averaged power and the mean blood pressure at control were 1.00±0.62 W and 92±9.3 mm Hg, respectively. The averaged maximum increase in power with nitroprusside was 22%, to 1.22±0.73 W, and this occurred at a mean pressure of 80±8.8 mm Hg. This study showed that the power-afterload relation in the human ventricle has a maximum power point at some intermediate level of afterload, similar to that found in animal studies. Instead of operating at the maximum power point, the failing ventricle operates far to the right of the peak because of inappropriately elevated afterload. This mismatch between the left ventricle and the arterial circulation resulted in a submaximal transfer of power. Afterload reduction improves power output by an improvement in matching between the left ventricle and the periphery. (Circulation Research1992;70:530–535)