This study was undertaken to compare the relative values of the low level predischarge exercise test and the postdischarge (6 weeks) symptom-limited test in 518 consecutive patients admitted with an acute myocardial infarction. Of the patients who did not develop significant ST segment depression or angina during the predischarge test, the symptom-limited test also remained negative in 91.5 and 91.9% of the patients, respectively. Similar results were obtained with ST segment elevation and the systolic blood pressure response during the two exercise tests with only 2.1 and 11.4% changing from normal to abnormal, respectively. Discriminant function analysis was done to predict the occurrence of coronary events (unstable angina, reinfarction, cardiac failure, cardiac death) with use of the data from the exercise tests together with other clinical and investigational data. The jackknife method correctly classified 71.9 and 71.4% of the patients with the data from the predischarge exercise test and symptom-limited test, respectively. Combining the data from the two tests improved the overall predictive accuracy to only 75.0%. It is concluded that the routine performance of a symptom-limited test 6 to 8 weeks after infarction does not reveal any significant additional information in those patients who have undergone a predischarge low level exercise test. Thus the 6 to 8 week test should be restricted to selected patients after myocardial infarction.