In patients who survive the acute phase of myocardial infarction, those with multivessel coronary artery disease generally have a worse prognosis than those with single-vessel disease. However, some patients with significant multivessel stenoses have a good prognosis, whereas some with a significant single-vessel stenosis have a poor prognosis. Thus, although definition of coronary anatomy may be helpful, it is a not a fail-safe prognosticator. In this retrospective analysis, the association of abnormalities at rest and during submaximal exercise testing with radionuclide ventriculography after acute myocardial infarction with major cardiac complications (death, recurrent infarction, severe angina or congestive heart failure) in the ensuing 6 months was assessed in patients with single and multivessel disease. Coronary angiography and submaximal exercise testing with radionuclide ventriculography were performed within 3 months of each other in 42 patients. Eleven of the 16 patients with single-vessel coronary stenosis had major cardiac complications. The subsequent course of these 16 patients was correctly predicted by left ventricular ejection fraction (LVEF) less than or equal to 0.40 in 8 patients, by LVEF less than 0.55 in 7 patients, by failure of LVEF to increase by 0.05 units in 13 patients, and by an increase in left ventricular end-systolic volume index (LVESVI) during exercise greater than 5% above baseline in 11 patients. Of the 26 patients with multivessel coronary artery disease, 24 had major cardiac complications. The subsequent course of these 26 patients was correctly predicted in 13 by LVEF less than or equal to 0.40, in 20 by LVEF less than 0.55, in 25 by a failure of LVEF to increase by 0.05 units during exercise, and in 20 by an increase in LVESVI by greater than 5% during exercise. Thus, submaximal exercise testing with radionuclide ventriculography may provide valuable prognostic information concerning the occurrence of major cardiac events after myocardial infarction not only in patients with multivessel disease, but also in those with single-vessel disease. Exercise-induced abnormalities of left ventricular function may have greater prognostic importance than the delineation of coronary arterial anatomy or the assessment of residual left ventricular function at rest.