OBJECTIVE: To assess the prevalence of domestic violence in pregnancy when midwives are trained to enquire about it routinely. DESIGN: A cross sectional study during a period after midwives had been trained to routinely enquire about it and a retrospective case note survey at an earlier period. SETTING: The maternity services of Guy's and St Thomas' NHS Hospital Trust in South London. SAMPLE: Women aged 16 and over booking for maternity care between 14th September 1998 and 21st January 1999. METHODS: Midwives were required to routinely enquire about domestic violence at booking, 34 weeks of gestation and postpartum (within 10 days) using a series of structured questions. MAIN OUTCOME MEASURES: The lifetime and annual rates of domestic violence. The prevalence of domestic violence in pregnancy. RESULTS: The prevalence of domestic violence in pregnancy was 1.8% at booking, 5.8% at 34 weeks of gestation and 5.0% at 10 days postpartum. Eight hundred and ninety-two women were asked about domestic violence on at least one occasion, of whom 22 (2.5%) reported domestic violence in pregnancy. Two hundred and sixty-five maternity notes were reviewed for the retrospective case note survey and one (0.37%) case of domestic violence in pregnancy was identified. Routine questioning increased the rate of detection of domestic violence by 2.1% (95% CI = 0.1-3.4%; P= 0.03). The lifetime prevalence of domestic violence was 13%, and 6.4% in the previous 12 months. CONCLUSIONS: Routine enquiry for domestic violence can increase the rate of detection in maternity settings, thereby providing an opportunity for women to access help early.