We evaluated the survival rates of medically and surgically treated patients with one-, two- or three-vessel coronary disease (CAD), stable pain, and normal left ventricular function observed at the Duke University Medical Center over an 11-year period. There was no statistical difference in survival when medical and surgical treatment were compared in patients with one-, two- or three-vessel disease (greater than or equal to 75% vessel occlusion). However, the survival curve of patients with three-vessel disease treated surgically consistently exceeded that of those treated medically over a 7-year follow-up period. This difference was not statistically significant, but suggests the need for further study in this group. When patients with 50% or greater lesions in three vessels, stable angina and normal left ventricular function were evaluated, surgical survival was greater than medical survival, although the difference is not statistically significant. The survival in the Duke medical group is substantially higher than that reported for medical patients in the European Cooperative Trial, which suggests that the significant differences surgical in survival in the European Cooperative Trial were largely due to a lower medical survival than that in the Duke medical group.