Cardiac peak power, a contractility index based upon instantaneous changes in intracavitary pressure and systolic peak flow, was measured at rest and during supine exercise in 26 patients with coronary artery disease and 8 healthy subjects. The pathophysiological significance of this index was compared with left ventricular ejection fraction (LVEF) during exercise. Cardiac peak power, ejection fraction, end-diastolic volume, stroke volume, cardiac output and systemic vascular resistance were measured at rest, during three stages of supine bicycle ergometry and two stages of recovery. Cardiac peak power increased continuously in healthy subjects, from 5.4 +/- 0.8 W/ml at rest to 11.4 +/- 3.1 W/ml at peak exercise, p < 0.001. In patients, peak power increased initially, reached a plateau in stage 2, and subsequently remained unchanged in stage 3 (5.6 +/- 2 versus 5.6 +/- 1.6 W/ml, p = ns). Ejection fraction demonstrated a flat response during exercise in patients, contrasting with a 42% increase in cardiac peak power. The lack of increase in ejection fraction was attributed to its dependence on afterload. Peak power showed no correlation with systemic vascular resistance (r = 0.01, p = ns). In a subgroup of patients with low resting LVEF (LVEF = 26% +/- 7%), peak power increased 70% during exercise, from 2.0 +/- 0.7 to 3.5 +/- 1.7 W/ml, p < 0.05, in contrast to a flat ejection fraction response. Thus, cardiac peak power, a relatively afterload-independent index of left ventricular performance and contractility can be obtained noninvasively during exercise.