Patients with postinfarction angina undergoing surgery for unstable angina face an increased risk of operative mortality. Between January 1982 and December 1987, clinical, angiographic, and operative data was collected prospectively in 588 unstable patients with a prior myocardial infarction within 30 days of surgery (MI) and 5951 unstable patients without preoperative damage (NONMI). MI patients were characterized as being older (age greater than or equal to 70 years: MI, 19.7%; NONMI, 11.6%; p less than 0.001) and having more left ventricular dysfunction (left ventricular ejection fraction less than 40%: MI, 34.8%; NONMI, 26.4%; p less than 0.001). Semi-elective surgery was performed in 82.0% of NONMI patients while 76.9% of MI patients underwent urgent surgery. Operative mortality was increased in MI patients (MI, 11.1%; NONMI, 4.0%; p less than 0.001) which was related to the extent of preoperative MI (non-Q wave, 8.3%; Q wave, 17.5%; p less than 0.001). Stepwise logistic regression analysis identified preoperative MI as an independent risk variable of operative mortality for unstable angina. Separate multivariate analyses were performed to identify the independent predictors for MI and NONMI patients. The multivariate predictors of operative death for MI patients were left ventricular dysfunction, reoperative coronary surgery, nonuse of the internal mammary, age, transmural MI (relative risk 2.11 vs non-Q wave infarction) and left main stenosis. For NONMI patients, the independent variables were urgent operation, left ventricular dysfunction, reoperation, female gender, left main stenosis, and age. The results of this study indicate that recent preoperative MI adversely influences the surgical results in patients with unstable angina. Alternative treatment strategies are warranted for high risk patients, particularly those with transmural MIs and impaired ventricular function.