The current status of the Veterans Administration Cooperative Study of the effect of surgery on survival in patients with stable angina is presented. The outcome in 686 adult males randomly allocated to medical or surgical treatment groups in 1972-1974 was studied in subgroups of patients classified by invasive (arteriographic) and noninvasive risk factors. In 91 patients with left main lesions reducing the luminal diameter 50% or more, surgery significantly improved survival in the two-thirds characterized as middle or high risk by four simple noninvasive predictors of prognosis (New York Heart Association functional classification III or IV, history of myocardial infarction, history of hypertension, and ST-segment depression on the resting baseline ECG as assessed on a centralized reading). Patients with three-vessel disease and no significant disease of the left main coronary artery also had better survival rates when treated surgically. However, this was statistically significant at 6 years only in the 10 hospitals in which the aggregate operative mortality was 3.3%. Patients without left main lesions were also categorized by four noninvasive predictors of risk. Categorizing such patients into roughly equal groups of high, middle, and low risk identified a high-risk group, in which surgery was associated with statistically improved survival, and low- and middle-risk groups in which it was not. The use of both invasive and noninvasive factors to assess risk in patients with chronic stable angina pectoris provided greater predictive power than either angiography or noninvasive factors alone.