While circadian variations in birth and perinatal mortality rates have previously been described in the literature, the reasons behind these observed rhythms remain unclear. The principal hypothetical causes include variations in obstetric practices and an association between the time of birth and biological parameters. In order to explore this issue we analysed the distribution patterns for time and day of birth, as well as circadian variations in maternal characteristics, obstetric practices and neonatal risk in a population at low obstetric risk. The study population included 685 low-risk pregnant women consecutively admitted at an early stage of labour to six maternity units. The results showed hourly variations in the birth rate and circadian variations in obstetric practices that might explain the hourly pattern observed for the birth rate. By contrast, the frequency of a positive neonatal risk indicator was uniform across all time categories in this population at low obstetric risk.