We reviewed blood use in 118 consecutive patients who underwent primary, elective cardiac operations in 1989. In June 1989 we initiated a blood conservation program that included attempts to limit preoperative aspirin use, intraoperative phlebotomy and hemodilution, use of a cell conservation device (Electromedics, Inc., Englewood, Colo.) to concentrate residual oxygenator contents, reinfusion of chest drainage, and acceptance of a minimum hemoglobin level of 8.0 gm/dl in stable patients. Patient characteristics were similar for patients operated on both before (n = 58) and after (n = 60) initiation of the blood conservation program, except for age and preoperative aspirin use (both greater in postconservation patients). Fewer blood products were transfused (5.8 +/- 5.7 units per patient before conservation versus 4.0 +/- 7.4 units per patient after conservation; p = 0.005). More complete data were available for 82 patients (40 patients before conservation and 42 after conservation). In the postconservation patients, 20 of 42 had 575 +/- 140 ml of blood withdrawn before cardiopulmonary bypass and reinfused afterward, 26 of 42 had 806 +/- 376 ml of blood processed with the cell conservation device returned, and 21 of 42 patients had an average of 287 +/- 127 ml of chest drainage reinfused. Chest tube drainage, postoperative hematologic parameters, and the prevalence of complications were not significantly different between groups. Stepwise linear regression analysis identified intraoperative withdrawal of blood before cardiopulmonary bypass, bypass duration, and preoperative hematocrit value as predictors of blood use. Intraoperative withdrawal of blood before cardiopulmonary bypass is an important conservation measure, and its use should be expanded.