Patients with postoperative myocardial infarction are frequently asymptomatic. Several follow-up strategies have been used to detect infarction or ischemia in asymptomatic patients. Different investigators have used quite different criteria to define patients at high risk. This study was done to evaluate these different approaches to selecting patients who should be monitored with electrocardiograms (ECG) or enzymes, or both, postoperatively, as well as different strategies for the timing of follow-up evaluation. A total of 232 patients, mostly hypertensive or diabetic, were evaluated before undergoing elective operations and were observed serially from the day of the operation until discharge or the sixth postoperative day with daily clinical evaluations, ECG and creatine kinase isoenzymes. Several follow-up strategies used in recent studies were evaluated for sensitivity and specificity in identifying the patient who had postoperative infarctions or ischemia. The most sensitive strategies would obtain ECG in asymptomatic patients on the day of the operation and the first two postoperative days. Several criteria for defining a high risk population were evaluated, including type of operation, age, history of cardiac disease, Goldman's cardiac risk classification and the results of the preoperative ECG. Monitoring of patients with an abnormal preoperative ECG would have identified 88 per cent of the patients with postoperative myocardial infarction and 63 per cent of the patients with definite ischemia. Goldman's risk class identified patients with a normal ECG who were at higher risk for postoperative ischemia.