Submaximal exercise testing with radionuclide ventriculography was performed in 117 patients prior to hospital discharge 16.7 +/- 6.7 days (SD) following acute myocardial infarction. The hypothesis tested in this study was that patients with different locations and types of infarction have different functional responses to submaximal exercise prior to discharge. The distribution of the myocardial infarctions were anterior transmural in 33, inferior transmural in 39, anterior nontransmural in 23, inferior nontransmural in 19, and indeterminant in three. Patients with transmural infarction generally had significantly larger resting left ventricular volumes at end-diastole and end-systole and lower ejection fractions and systolic blood pressure/end-systolic volume indexes than patients with nontransmural infarctions (p less than 0.05). During submaximal exercise, the change in end-systolic volume was significantly different in these two groups. When patients were separated further into anterior and inferior transmural subgroups, the patients with anterior transmural infarction had significantly lower left ventricular ejection fractions and higher right ventricular ejection fractions than the group with inferior transmural infarction (p less than 0.05). In response to exercise, the group with anterior transmural infarction had a significant decrease in left ventricular ejection fraction and a blunted systolic blood pressure/left ventricular end-systolic volume index, in comparison to patients with inferior myocardial infarction (p less than 0.05); this was the only group to have a significant increase in end-systolic volume. The group variance for the parameters studied was large, particularly during exercise when the individual responses were frequently directionally opposite from the group means. The group with anterior transmural infarction was the most homogeneous, with 26 of 33 having a directionally abnormal response to submaximal exercise. It was concluded that the group with anterior transmural infarction generally displayed the most abnormal left ventricular function. However, despite significant group differences in resting ventricular function with different infarcts, the intragroup variability at rest and in response to exercise was too great to permit an accurate prediction of the subject's resting ventricular performance or to permit a prediction of exercise response based solely on location of the infarct.