The measurement of left ventricular ejection fraction (LVEF) plays a key role in many strategies for managing patients after acute myocardial infarction. We tested the hypothesis that exercise capacity 1 month after myocardial infarction provides additional information in patients with a low LVEF and therefore assists in risk stratification. One hundred fifteen patients, with documented myocardial infarction and LVEF less than 35% by gated radionuclide scan 1 month after acute myocardial infarction, were followed up for 2 months to 7 years. Exercise capacity was estimated from a treadmill test 1 month after infarction. Using the Cox proportional hazards model, exercise capacity was a significant predictor of death or reinfarction. The relative risk of death, based on a comparison between the lowermost quintile (less than 4 METS) and uppermost quintile (greater than 7 METS), was 3.5 (95% confidence interval, 1.1-9.7); the relative risk in the fourth, third, and second quintile was 2.7, 2.1, and 1.6, respectively. In a multivariate analysis, the observed effect of a good exercise capacity was independent of LVEF. These data indicate that in patients with a low LVEF after myocardial infarction, useful prognostic information can be obtained from exercise testing.