This prospective study of symptom-limited supine ergometry was conducted to determine the contributions of right ventricular (RV) and left ventricular (LV) systolic function to the exercise capacity of a cohort of patients with coronary artery disease (CAD). Patients with unstable angina, angiographically proven CAD (n = 53) and stable symptoms after medical therapy or angioplasty were included. Documented myocardial infarction (greater than or equal to 2 weeks before exercise) was present in 43 of 53 patients. Angina was the limiting symptom in 11 of 53; the other 42 stopped exercise with dyspnea or fatigue, or both. Oxygen consumption was measured on-line during exercise with a metabolic cart. RV ejection fraction and LV ejection fraction were measured by validated methods from gated blood pool radionuclide ventriculography. There were weak but statistically significant correlations between exercise oxygen consumption and exercise RV ejection fraction (r = 0.30, p less than 0.05) and between exercise oxygen consumption and exercise LV ejection fraction (r = 0.38, p less than 0.01). Multivariate regression analysis, including exercise RV ejection fraction, exercise LV ejection fraction and exercise heart rate versus exercise oxygen consumption revealed a better relation (r = 0.48, p less than 0.005) than any variable in univariate regression. The values of RV and LV ejection fraction at rest did not correlate significantly (r = 0.2, difference not significant), but the exercise values did correlate weakly (r = 0.41, p less than 0.01). The reserve of LV ejection fraction, defined as exercise minus rest value, correlated weakly with exercise oxygen consumption (r = 0.32, p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)