Our objective was to determine the incidence of blood administration after cesarean delivery and whether such transfusions are always beneficial. In this retrospective study 1610 women underwent cesarean delivery during a 2-year period and 127 of these patients had hemorrhage during or after operation. Of these subjects 103 received blood because of intraoperative hemorrhage, a reduction in the hematocrit of more than 10 points, or because the postoperative hematocrit was < 24%. These subjects were compared with the remaining women (n = 24) who met the same criteria for hemorrhage but did not receive transfusion. The maternal age, race, and parity were similar in both groups. The estimated blood loss (873 +/- 484 ml) and the preoperative hematocrit (33.4% +/- 6.4%) in the women who did not receive transfusion were not significantly different from those of patients who received packed red blood cells (854 +/- 576 ml and 30.0% +/- 5.4%, respectively). The postdelivery hematocrit was similar in both groups: 25.9% +/- 4.3% in the nontransfused group and 24.5% +/- 5.6% in the transfused group. Patients in the transfused group received a mean of 3.8 +/- 4.9 units of packed red blood cells, with a range of 1 to 40 units. The mean equilibrated (stable) hematocrit after transfusion was 28.4% +/- 5.4%, which was significantly greater than the mean equilibrated postoperative hematocrit of 22.7% +/- 4.6% in patients who did not receive transfusion (p < 0.0001). Nonetheless, the hospital stay, incidence of postoperative infection, and incidence of wound complications were similar in the two groups.(ABSTRACT TRUNCATED AT 250 WORDS)