Objective: to study whether an increase in intrapartum referrals in primary midwife-led care births in the Netherlands is accompanied by an increase in caesarean sections. Design: nationwide descriptive study. Setting: The Netherlands Perinatal Registry. Participants: 807,437 births of nine year cohorts of women with low risk pregnancies in primary midwife-led care at the onset of labour between 2000 and 2008. Measurements: primary outcome is the caesarean section rate. Vaginal instrumental childbirth, augmentation with oxytocin, and pharmacological pain relief are secondary outcomes. Trends in outcomes are described. We used logistic regression to explore whether changes in the planned place of birth and other maternal characteristics were associated with the caesarean section rate. Findings: the caesarean section rate increased from 6.2 to 8.3 per cent for nulliparous and from 0.8 to 1.1 per cent for multiparous women. After controlling for maternal characteristics the year by year increase in the caesarean section rate was still significant for nulliparous women (adj OR 1.03; 95% CI 1.02-1.03). The vaginal instrumental birth declined from 18.2 to 17.4 per cent for nulliparous women (multiparous women: 1.7-1.5 per cent). Augmentation of labour and/or pharmacological pain relief increased from 23.1 to 38.1 per cent for nulliparous women and from 5.4 to 9.6 per cent for multiparous women. Conclusion: the rise in augmentation of labour, pharmacological pain relief and electronic fetal monitoring in the period 2000-2008 among women in primary midwife-led care was accompanied by an increase in caesarean section rate for nulliparous women. The vaginal instrumental deliveries declined for both nulliparous and multiparous women. Implications for practice: primary care midwives should evaluate whether they can strengthen the opportunities for nulliparous women to achieve a physiological birth, without augmentation or pharmacological pain relief. If such interventions are considered necessary to achieve a spontaneous vaginal birth, the current disadvantage of discontinuity of care should be reduced. In a more integrated care system, women could receive continuous care and support from their own primary care midwife, as long as only supportive interventions are needed.