To determine the long-term (2-year) cardiac prognosis of high-risk patients undergoing noncardiac surgery and to determine the predictors of long-term adverse cardiac outcome. Prospective cohort study. Historical, clinical, and laboratory data were collected during the in-hospital period, and at 6 months, 1 year, and 2 years following surgery. Data were analyzed using proportional hazards models. University-affiliated Veterans Affairs medical center. A consecutive sample of 444 patients with or at high risk for coronary artery disease who had undergone elective noncardiac surgery and were discharged from the hospital in stable condition. Cardiac death, myocardial infarction, unstable angina, progressive angina requiring coronary artery bypass graft surgery or coronary angioplasty, and new unstable angina requiring hospitalization. Forty-seven patients (11%) had major cardiovascular complications during a 728-day (median) follow-up period: 24 had cardiac death; 11, nonfatal myocardial infarction; six, progressive angina requiring coronary artery bypass graft surgery or coronary angioplasty; and six, new unstable angina requiring hospitalization. Thirty percent of outcomes occurred within 6 months of surgery and 64% within 1 year. Five independent predictors of long-term outcome were identified. Three predictors reflected the preexisting chronic disease state: (1) the presence of known vascular disease (hazard ratio, 6.1; 95% confidence interval [CI], 2.5 to 15.0; P less than .0001); (2) a history of congestive heart failure (hazard ratio, 5.0; 95% CI, 2.0 to 12.0; P less than .0005); and (3) known coronary artery disease (hazard ratio, 3.7; 95% CI, 1.7 to 8.0; P less than .0007). Two predictors reflected acute postoperative ischemic events: (1) myocardial infarction/unstable angina (hazard ratio, 20; 95% CI, 7.5 to 53.0; P less than .0001) and (2) myocardial ischemia (hazard ratio, 2.2; 95% CI, 1.1 to 4.3; P less than .03). Patients surviving a postoperative in-hospital myocardial infarction had a 28-fold increase in the rate of subsequent cardiac complications within 6 months following surgery, a 15-fold increase within 1 year, and a 14-fold increase within 2 years (95% CI, 5.8 to 32; P less than .00001). Seventy percent of all long-term adverse outcomes were preceded by in-hospital postoperative ischemia that occurred at least 30 days (median, 282 days) before the long-term event. The development of congestive heart failure or ventricular tachycardia (without ischemia) during hospitalization was not associated with adverse long-term outcome. The incidence of long-term adverse cardiac outcomes following noncardiac surgery is substantial. At increased risk are patients with chronic cardiovascular disease; at highest risk are patients with acute perioperative ischemic events. We conclude that survivors of in-hospital perioperative ischemic events, specifically myocardial infarction, unstable angina, and postoperative ischemia, warrant more aggressive long-term follow-up and treatment than is currently practiced.