Although cardiac surgery is performed in approximately 10% of acute myocardial infarction (AMI) patients undergoing a primary percutaneous transluminal coronary angioplasty (PTCA) reperfusion strategy before discharge, the indications for and timing of operative revascularization, and the short- and long-term outcomes after surgery have not been characterized. In the prospective, controlled Primary Angioplasty in Myocardial Infarction-2 trial, cardiac catheterization was performed in 1,100 patients within 12 hours of onset of AMI at 34 centers, followed by primary PTCA when appropriate. Cardiac surgery was performed before hospital discharge in 120 patients (10.9%), electively in 42.6%, and on an urgent or emergent basis in 57.4%. Surgery was performed in 6.1% of 982 patients after primary PTCA (although emergently for failed PTCA in only 4 cases [0.4%]), and in 53 of 118 patients (44.9%) not undergoing primary PTCA. Patients requiring surgery were older, and more frequently had diabetes and 3-vessel disease than those managed nonoperatively. Internal mammary artery grafts were placed in only 31% of patients. In-hospital mortality was 6.4% in patients undergoing urgent/emergent surgery, 2.0% after elective surgery, and 2.6% in patients not undergoing surgery (p = NS). After multivariate correction for baseline risk factors, early and late survival free of reinfarction were similar in patients undergoing versus not undergoing in-hospital cardiac surgery. Thus, the appropriate use of coronary artery bypass graft surgery in the peri-infarction period is an integral component of the primary PTCA approach, and is frequently used to optimize the prognosis of a high-risk AMI cohort with unfavorable baseline features. The implications for the performance of primary PTCA in AMI at centers without on-site surgical facilities are discussed.