The predictive value of radionuclide ventriculography was studied in 34 patients with depressed left ventricular ejection fraction (less than 40%) and clinically evident congestive heart failure secondary to atherosclerotic coronary artery disease. In addition to left ventricular ejection fraction, right ventricular ejection fraction and extent of left ventricular paradox were obtained in an attempt to identify a subgroup at increased risk of mortality during the ensuing months. The 16 patients who were alive after a 2 year follow-up period had a higher right ventricular ejection fraction and less extensive left ventricular dyskinesia. When a right ventricular ejection fraction of less than 35% was used as a discriminant, mortality was significantly greater among the 21 patients with a depressed right ventricular ejection fraction (71 versus 23%), a finding confirmed by a life table analysis. Depressed right ventricular function was further linked to more severely compromised left ventricular function, as confirmed by a greater reduction in left ventricular ejection fraction and by an increased extent of left ventricular dyskinesia. These patients had a greater prevalence of chronic obstructive pulmonary disease and previous inferior myocardial infarction but the differences between groups were not statistically significant. It appears that the multiple factors contributing to the reduction in right ventricular ejection fraction make it a useful index not only for assessing biventricular function, but also for predicting patient outcome.