The incidence of perioperative myocardial infarction with noncardiac surgery varies by the type of procedure and the prevalence of coronary atherosclerosis in the study population. Incidence is < or = 1% with minor procedures and may exceed 10% with vascular operations. The case fatality rate continues to be 30% to 50%. Pathogenesis is not understood completely. Diagnosis is sometimes problematic, because less than 50% of patients complain of chest pain. In addition, a high frequency of notable but apparently innocent postoperative electrocardiograph changes limits the diagnostic use of the electrocardiogram. Fortunately, the creatine kinase MB isoenzyme retains its sensitivity and specificity for acute infarction in perioperative patients. Different approaches to preoperative risk assessment have been developed, including a summative cardiac risk index and a stratification system based on the likelihood that the most powerful risk factor (coronary artery disease) is present. Although many interventions have been recommended to lower perceived risk, none has been tested in a randomized controlled trial, and their comparative efficacy and safety is unknown.