After myocardial infarction (MI), the additive prognostic value of exercise variables to clinical variables has been questioned. The merits of a symptom-limited predischarge exercise test were therefore evaluated in clinically defined subgroups of patients. Exercise tests were consecutively performed by 208 survivors of uncomplicated MI (no heart failure, postinfarction angina, recurrent infarction, or late arrhythmias) and by 92 survivors of complicated MI. After uncomplicated MI (1-year mortality rate 4%), an achieved workload greater than 70% of age-predicted maximum identified 145 patients at very low risk (predictive value for survival 98%). After complicated MI (1-year mortality rate 13%), an exaggerated heart rate response was the best predictor of outcome, but had low (92%) predictive value of survival at 155 bpm. It is concluded that stress testing has only limited value after complicated MI. After uncomplicated MI, exercise variables are extremely helpful in identifying patients at very low risk in whom further investigations are not warranted.