OBJECTIVE To compare the efficacy and safety of intravaginal and oral misoprostol vs. oxytocin/prostaglandin E2 (PGE2) gel for third trimester labor induction. METHODS Two hundred twenty-four pregnant women were randomized to induction of labor either with misoprostol or oxytocin and PGE2 gel. Patients in the misoprostol group (n = 112) received 100 micrograms intravaginal misoprostol followed by 100 micrograms p.o. every 2 h. The oxytocin/PGE2 group consisted of 112 patients who underwent PGE2 cervical instillation 6 h before continuous oxytocin infusion. The perinatal, intrapartum and neonatal characteristics of both groups were determined. RESULTS Induction to active phase of labor was successfully achieved in 96 women (85.7%) in the misoprostol group vs. 86 women (76.8%) in the oxytocin/PGE2 group, but the drug initiation-delivery interval was significantly shorter in the misoprostol group (9.2 +/- 2.4 h) than in the oxytocin/PGE2 group (15.2 +/- 3.2 h, P < 0.001). The incidence of adverse intrapartum outcomes was similar for both methods. Intravaginal misoprostol 100 micrograms followed by a single oral dose of 100 micrograms misoprostol safely produced labor and a vaginal delivery in 70% of patients. More than three tablets were required in only 10% of patients. There was a higher prevalence of cesarean section for failed induction in the oxytocin/PGE2 group than in the misoprostol group (13.4 vs. 6.3%, P < 0.001). The neonatal outcomes of both groups were also similar. CONCLUSIONS Misoprostol is significantly more effective for labor induction than oxytocin/PGE2 gel. The maternal intrapartum and neonatal outcomes were the same for both induction regimens. From a clinical and perinatal perspective, misoprostol is an acceptable choice for labor induction.