Objective: To determine if intraoperative glove change and placental delivery method affect the post-cesarean endometritis rate. Methods: After informed consent, women who required cesarean were randomly assigned to one of four study groups: 1) no glove change plus manual placental extraction, 2) no glove change plus spontaneous placental delivery, 3) glove change plus manual extraction, and 4) glove change plus spontaneous delivery. Bilateral glove change by both primary and assistant surgeons occurred immediately after delivery of the newborn and before delivery of the placenta. External uterine massage and traction on the umbilical cord were performed to assist spontaneous delivery of the placenta. A first-generation cephalosporin was routinely administered after umbilical cord clamping for prophylaxis of post- cesarean endometritis. Results: Of 760 women entered into the study, we included 643 who did not have intrapartum chorioamnionitis or cesarean hysterectomy. The four groups were comparable with respect to selected maternal and intrapartum characteristics, including maternal and gestational age, parity, presence of labor, and the presence and duration of membrane rupture. The postoperative endometritis rate was significantly higher in women whose placentas were extracted manually (31 versus 22%, P = .01). Operator glove change did not alter the incidence of endometritis (relative risk 1.0, 95% confidence interval 0.79-1.3). Conclusion: Manual extraction of the placenta is associated with a significantly greater risk of post-cesarean endometritis than that observed with assisted spontaneous placental delivery. Intraoperative glove change does not decrease post-cesarean endometritis.