The association between leisure-time physical activity (LTPA) and the risk of sudden cardiac death (SCD) in coronary artery disease patients is not well known. We aim to assess whether there is an association between LTPA and the risk of SCD and non-SCD in patients with coronary artery disease. Patients with angiographically verified coronary artery disease (n=1946) underwent a clinical evaluation, including filling in an LTPA questionnaire and extensive risk profiling at the baseline. The patients were classified into 4 groups according to LTPA: (1) inactive; (2) irregularly active; (3) active, exercise regularly 2× to 3× weekly; (4) highly active, exercise regularly ≥4× weekly. Age, sex, body mass index, left ventricular ejection fraction, type 2 diabetes mellitus, history of myocardial infarction, Canadian Cardiovascular Society grading of angina pectoris class, and exercise capacity were used as covariates in the multivariate Cox regression analysis. During follow-up (median 6.3 years), 52 SCDs and 49 non-SCDs occurred. Inactive patients had increased risk for SCD compared with active patients (hazard ratio, 2.45 [95% CI, 1.01-5.98]; P<0.05). A significant LTPA×Canadian Cardiovascular Society grading of angina pectoris class interaction was observed in SCD risk (P=0.019 in highly active patients). LTPA was not associated with SCD in patients with Canadian Cardiovascular Society grading of angina pectoris class 1 (n=1107, 18 events). Among patients with Canadian Cardiovascular Society grading of angina pectoris class 2 or higher (n=839, 34 events), increased risk for SCD was encountered in highly active patients (hazard ratio, 7.46 [95% CI, 2.32-23.9]; P<0.001) and inactive patients (hazard ratio, 3.64 [95% CI, 1.16-11.5]; P<0.05) as compared to active patients. A linear association was observed between LTPA and non-SCD; those with high LTPA had the lowest risk for non-SCD. Inactive coronary artery disease patients had increased risk for SCD. In subgroup analysis among symptomatic patients, the risk of SCD was increased in highly active and inactive patients compared with active patients. Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT01426685.