Submaximal exercise testing with radionuclide ventriculography (RVG) was performed in 117 patients before hospital discharge 17 +/- 7 days (+/- standard deviation) after an acute myocardial infarction (MI). The hypothesis tested in these studies was that submaximal exercise testing coupled to RVG allows the identification of patients at risk for future ischemic events in the subsequent 6 months, irrespective of MI location and type. The sites of MI were characterized as anterior transmural in 33, inferior transmural in 39, limited nontransmural in 18, extensive nontransmural in 24 and indeterminant in 3. During 6 months of follow-up, 9 patients died, 14 had recurrent MI, 18 had refractory angina pectoris, 16 had limiting angina and 17 had congestive heart failure. Discriminant function analysis ranked exercise changes in left ventricular (LV) ejection fraction and end-systolic volume the most important of all clinical, exercise and scintigraphic variables for predicting future cardiac events. The predictive accuracy of changes in LV ejection fraction and end-systolic volume were 93 and 91%, respectively, for the entire group, and were significantly more sensitive than any degree of ST-segment depression or elevation (p less than 0.001). These findings were generally independent of MI location and type. Thus, submaximal exercise RVG after MI is an accurate means of identifying patients at risk for major cardiac events in the 6 months after hospital discharge.