In February 1991 I initiated a blood conservation program that included hemodilution, use of a cell conservation device to concentrate diluted blood from operative field and from mediastinal shed blood, acceptance of a minimal hematocrit level of 25% in stable patients, and strict avoidance of use of blood products without definitive indication. A retrospective study to evaluate the effects of blood conservation program was designed to compare the amount of homologous transfusion, hematologic data, and postoperative outcome in patients operated on before and after initiation of blood conservation program. Patient characteristics were not different between two groups, before[n = 18] and after[n = 42] initiation of the program. The significant decrease of homologous transfusion[5.2 vs 1.4 units] and the marked increase of nontransfused patients[none vs 57%] in experimental group were due to marked decrease in homologous transfusion during the operation[4.6 vs 0.7 unit]. I conclude that with strict intraoperative blood conservation program, cardiac patients can be operated on with minimal homologous transfusion.