Chronic heart failure (CHF) studies investigating the clinical, hemodynamic, and therapeutic importance of endothelin-1 (ET-1), atrial natriuretic peptide (ANP), and brain natriuretic peptide (BNP) are largely based on resting plasma levels, which may vary with prior exertion and postprandial status. This study investigated the importance of peak-exercise plasma levels of ET-1, ANP, and BNP in the assessment of left ventricular (LV) systolic function. Thirty-six male-patients ages 58 ± 10 (mean ± SD) with NYHA class I-IV CHF due to coronary artery disease or idiopathic dilated cardiomyopathy were enrolled. LV systolic function was assessed by echocardiography and radionuclide ventriculography. Resting and peak cardiopulmonary exercise venous blood sampling and treadmill exercise testing were performed in the fasting state. Resting plasma levels of ET-1, ANP, and BNP were elevated compared with reference laboratory normal values. Exercise induced significant (p < 0.0001) increase in plasma levels of ET-1, ANP, and BNP. On univariate analysis peak-exercise plasma levels of ET-1, ANP, and BNP were more closely related to echocardiographically determined LV end-diastolic diameter and end-systolic diameter than their resting values. Multiple step-wise regression models identified resting and peak-exercise plasma levels of ET-1 and ANP but only the resting BNP as independent predictors of LV dimensions and systolic function. Peak exercise plasma levels of ANP and ET-1 are potentially more reliable and important than their resting levels as markers of LV systolic dysfunction and LV dimensions in patients with heart failure.