Self-reported leisure-time physical activity level correlates well with both cardiovascular (CV) and non-CV mortality in subjects without coronary heart disease (CHD). The impact of leisure-time physical activity on long-term outcomes has not been well studied in patients with preexisting CHD, who are often physically limited because of symptoms, medications, and co-morbid conditions. The aim was to determine the long-term prognostic value of self-reported leisure-time physical activity in a large CHD cohort. Leisure-time physical activity was evaluated using a self-administered questionnaire and categorized using a 4-level scale (sedentary, mild, moderate, and strenuous) in 14,021 of 24,958 subjects from the Coronary Artery Surgery Study Registry with suspected or proven CHD who underwent cardiac catheterization from 1974 to 1979. Median long-term follow-up was 14.7 years (interquartile range 9.8 to 16.2). Clinical outcomes were evaluated according to physical activity level and adjusted for potential confounders. Long-term all-cause and CV mortality progressively increased from most to least active subjects, with sedentary patients showing a 1.6-fold increase in mortality for both these outcomes (hazard ratio [HR] 1.63, 95% confidence interval [CI] 1.34 to 1.97, p <0.0001 for all-cause mortality). Similar trends were noted for men and women and in adjusted models, although HRs were attenuated after adjusting for age, gender, smoking, hypertension, diabetes mellitus, total cholesterol, body mass index, and ejection fraction (adjusted HR 1.23, 95% CI 1.01 to 1.49, p = 0.03 for all-cause mortality; adjusted HR 1.25, 95% CI 0.99 to 1.57, p = 0.05 for CV mortality). In conclusion, leisure-time physical activity independently predicted long-term survival in men and women with chronic stable CHD.