Previous comparisons of medical and surgical therapy for coronary artery disease were performed in the 1970s and may need to be updated to reflect current treatment efficacy. The purpose of this investigation was to determine the impact on long-term patient survival of changes over time in medical and surgical therapy. Among 5125 patients referred for catheterization between 1969 and 1984, 2261 underwent surgery, and 2864 received medical therapy. Survival in both medically and surgically treated patients improved over time. The improvement in medical patients was due to less sick patients being treated (a decreasing baseline risk), while that observed in surgical patients was due to refinements in care. The rate of improvement in survival over time was much greater in surgical patients, and the difference as compared with medical therapy was highly significant (treatment interaction with time p less than .0001). Estimates of 5 year survival adjusted for baseline risk factors for a patient in 1977 with one-, two-, or three-vessel disease and an ejection fraction of 0.40 were 0.88, 0.80, and 0.64 in medically treated, and 0.88, 0.87, and 0.80 in surgically treated patients. Corresponding projected estimates for 1984 were unchanged for medical patients but improved for surgical patients (0.93, 0.92, and 0.90). Thus, the projected survival benefits of surgery in 1984 appear more significant for all categories of coronary artery disease because of a differential improvement in surgical therapy. These data are relevant to therapeutic decision making for current patients and emphasize the importance of continued analysis of this topic as treatment efficacies change over time.