We assessed the effects of exercise testing, home exercise, and supervised exercise training on the physical-activity habits of patients after coronary artery bypass graft surgery. Sixty-five patients were assigned each to one of three groups: a control group receiving no exercise testing or exercise follow-up; a second group receiving exercise testing and exercise prescription followed by home exercise training; and a third group receiving exercise testing and prescription followed by gymnasium-based supervised exercise training one to three times a week. The exercise-trained groups underwent repeat exercise testing and prescription 12 weeks after discharge from the hospital. Six to eight months after surgery, all groups completed the Harvard Alumni Activity Survey. Physical activity among patients in the two exercise training groups was similar, whereas the control group reported significantly less leisure-time physical activity than the former two groups. The supervised-exercise-training group showed a significantly greater improvement in peak functional capacity than did the home-exercise-training group.