Objective To describe the outcome of night onset of labour as compared to the day onset of labour to investigate if labour which begins at night is more efficient.

Design Retrospective review of labour and delivery data.

Setting A large UK maternity service.

Population Over the period of 10 years 30 022 eligible women delivered, of which 19 842 deliveries were studied.

Methods A UK maternity department database was used to identify deliveries over a 10 year period and the delivery outcomes were retrieved from these records. 19 842 labours were divided in two categories; 11 794 night onset (2200–0600 h) and 8048 day onset (1000–1800 h) labours.

Main outcome measures Rates of caesarean section, artificial rupture of membranes (ARM) and syntocinon augmentation, epidural and labour duration.

Results A significant difference in delivery outcome was noted (p=0.004) with the night onset labours having more normal deliveries (83.6% vs 82.5%), fewer caesarean sections (8.7% vs 10.1%), fewer labour augmentations with syntocinon (14.9% vs 19.5%, p<0.001), fewer ARM (14.1% vs 15.6% P<0.001) and a significantly shorter mean first stage duration (4 h 58 min vs 5 h 7 min, p<0.05).

Conclusions Labours that start at night appear to be more efficient than labours that start during the day time.