ObjectiveTo assess whether labour variables (i.e. individuals characteristics, labour characteristics and medical interventions) impact maternal and newborn microbiomes.DesignProspective monocentric study.SettingSaint‐Joseph Hospital tertiary maternity unit, in Paris, France.PopulationAll consecutive primiparous women with a physiological pregnancy and term labour from 15 April to 1 June 2017.Methods16S ribosomal RNA gene sequencing of the maternal vaginal, newborn skin and newborn oral microbiomes from 58 mother–baby dyads.Main outcome measuresAnalysis of the effects of 19 labour variables on the composition and diversity of these microbiomes.ResultsThe 19 labour variables explained a significant part of the variability in the vaginal, newborn oral and skin microbiomes (44%–67%). Strikingly, duration of rupture of membranes was the single factor that explained the greatest variability (adjusted R2: 7.7%–8.4%, p ≤ 0.002) and conditioned, by itself, the compositions of the three microbiomes under study. Long duration of rupture of membranes was specifically associated with a lower relative abundance of the Lactobacillus genus (1.7‐fold to 68‐fold reduction, p < 0.0001) as well as an increase in microbiome diversity, including genera implicated in nosocomial infections. The effects of duration of rupture of membranes were also present in newborns delivered by non‐elective caesarean section.ConclusionsMaternal and newborn microbiomes were greatly affected by labour variables. Duration of rupture of membranes, even in non‐elective caesarean sections, should be considered in epidemiological and microbiological studies, as well as in vaginal seeding practices.