To compare the obstetric outcome of induction of labor at 41 weeks and of follow-up until 42 weeks and induction if the patient has still not given birth at 42 weeks. Six hundred women at 287+/-1 days of gestation with definitely unfavorable cervical scores were randomized to labor induction (N=300) or spontaneous follow-up (N=300) with twice-weekly nonstress testing and amniotic fluid measurement and once-weekly biophysical scoring. The treatments used in the induction group were (1) vaginal administration of 50 microg misoprostol (n=100), (2) oxytocin induction (n=100), and (3) transcervical insertion of a Foley balloon (n=100). The primary outcome measures were the cesarean delivery rate, whether or not the normal hospital stay had to be extended, and the neonatal outcomes. Secondary outcome measure included number of emergency cesarean deliveries performed for abnormalities of the fetal heart rate (FHR). The abdominal delivery rate was 19.3% in the induction group and 22% in the follow-up group (p=0.4). The mean length of hospital stay in the two main groups was 1.4+/-0.8 days and 1.3+/-1 days, respectively (p=0.1). Significantly higher rates of macrosomia and shoulder dystocia were seen in the follow-up group (24.6 and 2.3%) than in the induction group (7.6%, p<0.001; 0.3%, p=0.03). Meconium-stained amniotic fluid and meconium aspiration syndrome were observed significantly less frequently in the induction group (9.3 and 1.3%) than in the follow-up group (20.3%, p<0.001; 4%, p=0.03). Rates of emergency abdominal delivery in response to worrying FHR traces, neonatal intensive care unit admission, and low umblical artery pH were similar in the two groups. There was one intrauterine fetal death in the follow-up group. Induction of labor at 41 weeks of gestation does not increase the cesarean delivery rate or cause a longer stay in hospital than follow-up until 42 weeks, and neonatal morbidity is also lower after induction.