Objective: to test the hypothesis that midwives working in higher intervention units would have a higher perception of risk for the intrapartum care of women suitable for midwifery-led care than midwives working in lower intervention units. Methods: an initial retrospective analysis of the computerised records of 9887 healthy Caucasian women in spontaneous labour enabled the categorisation of 11 units as either ‘lower intrapartum intervention’ or ‘higher intrapartum intervention’ units. A survey of the midwives involved in intrapartum care in these 11 units, using standardised scenario questionnaires, was used to investigate midwives’ options for intrapartum interventions, their perceptions of intrapartum risk and the accuracy of these perceptions in the light of actual maternity outcomes. Findings: midwives working in maternity units that had a higher level of intervention generally perceived intrapartum risks to be higher than midwives working in lower intervention units. However, midwives generally underestimated the ability of women to progress normally and overestimated the advantages of technological interventions, in particular epidural analgesia. Conclusions: variations in intrapartum care cannot be solely explained by the characteristics of the women. The influence of the workplace culture plays a significant role in shaping midwives’ perceptions of risk, but it seems even more likely that the medicalisation of childbirth has had an influence on midwives’ appreciation of intrapartum risks. Intervention rates for low-risk births are often higher than recommended by research. The level of interventions varies across hospitals and higher rates are associated with higher perception of risk by midwives. Attention needs to be given to the influence the workplace plays in shaping midwives’ perception of risk; and to the effect of organisational culture on intervention rates.