The purpose of this study was to bring up-to-date the immediate risks and long-term benefits of coronary surgery in elderly people. The results obtained in a series of 475 patients in their seventies suffering from coronary disease who underwent myocardial revascularization alone or combined with heart valve replacement at the La Pitié Hospital, Paris, between 1984 and 1989 were analyzed. Coronary disease was associated with heart valve disease in two-thirds of the cases. The mean number of bypasses was 2.3 per patient. Aortic valve replacement was performed in 119 cases and mitral valve replacement in 12 cases. The operative mortality rate was 10.8 percent in patients with stable angina who underwent elective surgery. The factors which increased the mortality rate were: age (19.6 percent over 75 versus 13 percent under 75, P = 0.05); sex (23.5 percent in women versus 13.2 percent in men, P = 0.01) and emergency (25.8 percent, P = 0.02). No significant difference in mortality was noted between patients who had and those who did not have aortic valve replacement (15.8 versus 10.8 percent, P = 0.14). Fifteen patients (3.1 percent) developed perioperative myocardial infarction. During the follow-up, period myocardial infarction occurred in 7 patients. Out of 29 late deaths, 8 were of cardiac origin (infarction in 5, terminal heart failure in 3). The survival rate at 4 years was 76 percent. Among the patients who could be followed up, 80 percent are now asymptomatic, 15 percent are in stage I or II and 5 percent in stage III or IV of the Canadian Cardiovascular Society classification. Thus, despite a non negligible operative risk the functional improvement and survival rates obtained justify an increase in the indications for myocardial revascularization in elderly patients.