Objective To compare the cesarean rates in women undergoing induction of labor with unfavorable cervices who receive either misoprostol or extra-amniotic saline infusion. Methods We assigned 250 women undergoing indicated induction of labor randomly to misoprostol (50 μg every 4 hours for three doses with or without oxytocin) or extra-amniotic saline infusion and oxytocin. Each eligible subject had a singleton gestation, vertex presentation, intact membranes, cervical dilation no more than 2 cm, and effacement no more than 50%. Results Two hundred forty-eight women were studied after two exclusions; 120 were assigned to misoprostol, and 128 to extra-amniotic saline infusion. This sample size ensured an 80% chance of detecting an intergroup difference in cesarean rates of at least two-fold. The groups were similar in age, race, parity, indication for induction, and gestational age. The extra-amniotic saline infusion group had a significantly lower median initial dilation; however, at 12 hours, this group was significantly more dilated. The groups also were similar in epidural use, chorioamnionitis rates, endometritis rates, birth weights, and neonatal outcomes. Overall, 23% of women underwent cesarean deliveries, 18% and 27% for the misoprostol and extra-amniotic saline infusion groups, respectively (P = .12). There were no significant differences in the median time to either vaginal or cesarean delivery between the two groups. Eighty-one percent of women were delivered in less than 24 hours from initiation of induction; 78% in the misoprostol group, and 84% in the extra-amniotic saline infusion group (P = .19). Conclusion Both methods of labor induction appear to be equally effective in this group of women at high risk for cesarean delivery and prolonged induction of labor. © 1998 Lippincott Williams & Wilkins, Inc.