Objective Neonatal complications related to timing of elective cesarean section (ECS) have never been studied in randomized trials. We designed the first randomized trial of timing of ECS and hypothesized a decrease in neonatal admission rate if ECS was scheduled after 39 completed weeks of gestation. Material and methods Participants were randomized to either ECS performed at 38+3 weeks of gestation (± two days) or ECS performed at 39+3 weeks of gestation (± two days). Eligibility criteria were uncomplicated pregnancies with a single fetus and a reliable due date decided by ultrasound. Diabetics and women with an estimated high risk of having ECS before 39 weeks and 5 days of gestation were excluded. The primary outcome was admission to the Neonatal Intensive Care Unit within 48 hours of birth. Results From March 2009 to June 2011 1274 women from seven Danish hospitals were enrolled. Baseline characteristics were similar in intervention groups. No significant difference in primary outcome was found between ECS delivery at 38+3 weeks (88/635 neonates or 13.9% admitted) and ECS delivery at 39+3 weeks (76/637 neonates or 11.9% admitted), RR 0.86 (95% CI 0.65-1.15). Compliance was defined as ECS delivery date according to randomization or at any other date due to complications or labor, and numbers were 591/636 (93%) in group 38+3 and 567/638 (89%) in group 39+3, respectively. Conclusions ECS delivery scheduled at 39+3 weeks compared to 38+3 weeks did not significantly reduce neonatal admission rate. (Funded by The Danish Council for Independent Research and others; ClinicalTrials.gov number NCT00835003).