Rest and exercise radionuclide ventriculography were assessed for their value in predicting major cardiac events in patients with chest pain. Of 219 patients who were followed for up to 51 months, 42 had major cardiac events: 12 patients (5.5%) died, 5 (2.2%) sustained nonfatal myocardial infarction, and 25 (11.4%) had coronary arterial bypass grafting. Univariate and multivariate survival analysis revealed that exercise left ventricular ejection fraction was the best predictor for total major events and the resting ejection fraction to be the best predictor for death or nonfatal myocardial infarction. These two variables were strong predictors in the entire group of patients and in subgroups: patients with or without Q-wave infarction, patients with high probability of coronary artery disease and those with abnormal resting left ventricular function. Thus, radionuclide angiography provides important prognostic data that permits the physician to categorize patients with chest pain syndromes with respect to subsequent cardiac events. If validated, this model or a modification of it could identify patients at high risk of subsequent major cardiac events who are candidates for intensive follow up and therapy or further invasive evaluation, as well as patients at low risk of subsequent major cardiac events for whom standard follow up would be appropriate. © 1984.