To evaluate the role of analysis of right ventricular function with exercise in patients with presumed coronary artery disease referred for radionuclide ventriculography, the records of 55 patients referred to our laboratory over a 19-month period were reviewed. All underwent rest and exercise first-pass radionuclide stress testing and cardiac catheterization within a period of four months. Three groups were identified: (1) patients with normal exercise right ventricular function (n = 24); (2) patients with exercise-induced right ventricular regional wall motion abnormalities (n = 15); and, (3) patients with abnormal resting right ventricular function without new exercise abnormalities (n = 16). Patients in each group were similar in age, sex, baseline left ventricular function, medication usage, and indication for study. The incidence of right coronary artery disease was identical in the three groups, as was the incidence of left ventricular functional abnormalities with exercise. Patients with proximal right coronary artery disease were more likely to have reduced left ventricular ejection fraction and more extensive coronary artery disease than those without disease at this site. We conclude that: (1) analysis of rest and exercise right ventricular function does not allow prediction of coronary anatomy in an unselected group of patients; (2) normal right ventricular function with exercise is compatible with extensive coronary artery disease, including proximal right coronary artery disease; and (3) abnormal exercise right ventricular function may be due to exertional left ventricular dysfunction in the absence of proximal right coronary artery disease.