ObjectivesOur objective was to describe contemporary practice patterns in the timing of caesarean delivery in relation to cervical dilation, overall and by indication for caesarean. Our secondary objective was to examine how commonly caesarean delivery was performed for labour dystocia at dilations below 4 cm or without the use of oxytocin, overall and between hospitals.DesignRetrospective, population‐based cohort study.SettingOntario, Alberta, and British Columbia, Canada, 2008–2012.PopulationNulliparous women in labour who delivered term singletons in cephalic position.MethodsHistograms were used to examine the distribution of cervical dilation at time of caesarean delivery, overall and by indication for caesarean. Funnel plots were used to illustrate variation in hospital‐level rates of caesarean deliveries for labour dystocia that were performed early (<4 cm dilation) or without the use of oxytocin.Main outcome measuresCervical dilation (in centimetres) at time of caesarean delivery.ResultsThe population‐based cohort comprised 392 025 women, of whom 18.8% had a caesarean delivery. Of first‐stage caesareans for labour dystocia in women who entered labour spontaneously, 13.6% (95% CI 12.9, 14.2) had dilations <4 cm [hospital‐level inter‐quartile range (IQR): 6.2% to 20.0%] and 29.5% (95% CI 28.6, 30.4) did not receive oxytocin to treat their dystocia (hospital‐level IQR: 22.1–54.6%).ConclusionsThe proportion of caesareans done before 4 cm dilation or without oxytocin varies substantially across hospitals and suggests the need for institutions to review their practices and ensure that management of labour practice guidelines are followed.Tweetable abstractMany caesareans for labour dystocia are performed early during labour (<4 cm dilation) or without oxytocin.