Intra-aortic balloon pumping (IABP) to assist the failing circulation has become widely applied and accepted since its introduction in 1968. The elective, preoperative use of IABP for patients undergoing cardiac surgery has now become the controversy. The purposes of this report are to examine our experience with IABP and to determine its appropriate role in high-risk patients. IABP was utilized in 75 of 2333 (3.2%) adult cardiac surgical patients at Emory University Hospital from January 1976 through June 1978. IABP was required for refractory shock following cardiopulmonary bypass (CB) in 53 patients, for preoperative cardiogenic shock after acute myocardial infarction (CSMI) in nine and was electively placed prior to CB in 13. Sixty-two patients (81%) were able to separate from CB with IABP and pharmacologic support and were assisted 24–432 hours (median 64 hours). Fifty-five (73%) were weaned from IABP. Fifty (67%) are hospital survivors; late deaths have occurred in six patients (8%). Hemodynamic effect of IABP was demonstrated by comparison of pumping 1:1 to 1:8 mode in five balloon-dependent patients after CB. IABP was found to decrease systolic blood pressure, left ventricular filling pressure and peripheral resistance (p <.05). It increased diastolic and mean blood pressure, cardiac index and the endocardial viability ratio (p <.05). The post-CB use of IABP resulted in highest salvage when utilized to support failing hearts that required surgery despite recent preoperative infarction or when intraoperative ischemic injury had occurred. Poorest results were in patients with extensive chronic myocardial damage. Except in the case of preoperative cardiogenic shock, it was impossible to establish statistically reliable criteria for patients in whom elective preoperative insertion was found to be necessary. Careful surgical and anesthesia management with good monitoring can be used instead of preoperative IABP in the majority of (if not all) hemodynamically stable patients regardless of risk classification.