Background There is a continuing debate regarding optimal timing and mode of delivery in twins.

Aim To describe neonatal mortality rates (NMR) and factors influencing mode of delivery in twins following induced or spontaneous labour.

Methods We used data from the population-based Northern Survey of Twin and Multiple Pregnancy on twins delivered in the North of England during 1998–2007. Stillbirths and neonatal deaths were identified from the Northern Perinatal Mortality Survey. Factors associated with mode of delivery in twins with spontaneous and induced labour were explored by logistic regression.

Results There were 4203 twin pairs alive at the onset of labour or prior to planned caesarean section (CS) of which 29.7% (n=1250) underwent elective CS, 43.9% (n=1843) laboured spontaneously, and 20.6% (n=862) were induced. For spontaneous labours, the NMR was 40.3 and 44.0 per 1000 live births for first and second twins respectively (p=0.6), and for induced labours, 1.2 and 7.0 per 1000 live births respectively (p=0.12). Twins delivered following spontaneous labour were born at significantly earlier gestation than following induction: 34.2 (SD±3.7) versus 37.2 weeks (SD±1.4) (p<0.001). There was a higher rate of vaginal delivery of both twins following induced labour: 68.6% versus 63.3% following spontaneous labour (p<0.01). Presentation of both twins, maternal age and parity were associated with mode of delivery in both groups.

Conclusion There was a higher rate of vaginal delivery of both twins after induced compared with spontaneous labour. The higher NMR in twins after spontaneous labour was associated with earlier gestation at delivery.