The intra-aortic balloon pump (IABP) is an established additional support to pharmacologic treatment of the failing heart after myocardial infarction, unstable angina and cardiac surgery. The physiologic effect of IABP are: increased coronary perfusion and decreased left ventricular outflow impendance. The effects of preoperative IABP are not established. The aim of this study was to define indications and evaluate the effect of preoperative IABP support. Between January 1, 1990 and December 31, 1994, 1999 patients underwent CABG [318 REDOCABG, 15.9% and 1681 primary CABG, 84.9%]. From January to June 1994, 80 of 1811 patients (4.5%) required additional intra-aortic balloon pump (IABP) support postoperatively, 35 after REDO (21%) and 45 after primary CABG, 3%. Indication for IABP was failure to wean from cardiopulmonary bypass (CPB), despite maximal dose pharmacological support. The overall hospital mortality was 11.6% after REDO's and 2.1% after primary CABG, p < 0.001. In those where IABP was required the mortality rate was 67.7% after REDO and 36.8% after primary CABG, p < 0.01. Multivariate analysis identified preoperative unstable angina as independent risk factor for mortality (p < 0.001). REDO-CABG, LVEF < 40%, diffuse coronary artery disease and left main stem stenosis were identified as riskfactors for mortality when combined with one and another in a combination of at least two factors. In the IABP supported group mean ICU stay was 5.4 +/- 1.8 days. Thereafter the concept of preoperative use of IABP was introduced and presence of at least two of the above identified riskfactors served as indication for preoperative IABP treatment. The IABP was placed in the operating room at induction of anesthesia (1-1.5 hr prior to cross-clamping). Between June and December 1994, 19 of 188 patients received a preoperative IABP (9.0%). There were no balloon related complications. The mortality rate was 21%. Weaning from CPB was easy and in 30% of the cases the IABP was not necessary postoperative. In 90% of the pumps were removed within 24 hours postoperative. The mean ICU stay was 2.1 +/- 0.9 days. The first part of this study identified riskfactors for mortality after CABG and in the second series the efficacy of preoperative IABP treatment in "high-risk" patients was demonstrated. Preoperative IABP support in these patients seems also cost efficient due to lower mortality and shorter stay in intensive care unit in this group of patients.