The Coronary Artery Surgery Study (CASS) randomized 780 patients to an initial strategy of coronary surgery or medical therapy. Of medically randomized patients, 6% had surgery within 6 months and a total of 40% had surgery by 10 years. At 10 years, there was no difference in cumulative survival (medical, 79% vs. surgical, 82%; NS) and no difference in percentage free of death and nonfatal myocardial infarction (medical, 69% vs. surgical, 66%; NS). Patients with an ejection fraction of less than 0.50 exhibited a better survival with initial surgery treatment (medical, 61% vs. surgical, 79%; p = 0.01). Conversely, patients with an ejection fraction greater than or equal to 0.50 exhibited a higher proportion free of death and myocardial infarction with initial medical therapy (medical, 75% vs. surgical, 68%; p = 0.04) although long-term survival remained unaffected (medical, 84% vs. surgical, 83%; p = 0.75). There were no significant differences either in survival and freedom from nonfatal myocardial infarction, whether stratified on presence of heart failure, age, hypertension, or number of vessels diseased. Thus, 10-year follow-up results confirm earlier reports from CASS that patients with left ventricular dysfunction exhibit long-term benefit from an initial strategy of surgical treatment. Patients with mild stable angina and normal left ventricular function randomized to initial medical treatment (with an option for later surgery if symptoms progress) have survival equivalent to those patients randomized to initial surgery.