OBJECTIVE: To determine the association between ethnic group and risk of postpartum haemorrhage in women giving birth. DESIGN: Cohort study. SETTING: Maternity units in England. SAMPLE: A total of 981 801 records of births between 1 April 2015 and 31 March 2017 in a national clinical database. METHODS: Multivariable logistic regression analyses with multiple imputation to account for missing data and robust standard errors to account for clustering within hospitals. MAIN OUTCOME MEASURE: Postpartum haemorrhage of ≥1500 ml (PPH). RESULTS: A total of 28 268 (2.9%) births were complicated by PPH. Risks were higher in women from black (3.9%) and other (3.5%) ethnic backgrounds. Following adjustment for maternal and fetal characteristics, and care at birth, there was evidence of an increased risk of PPH in women from all ethnic minority groups, with the largest increase seen in black women (adjusted OR 1.54, 95% CI 1.45-1.63). The increase in risk was robust to sensitivity analyses, which included changing the outcome to PPH of ≥3000 ml. CONCLUSIONS: In England, women from ethnic minority backgrounds have an increased risk of PPH, when maternal, fetal and birth characteristics are taken into account. Factors contributing to this increased risk need further investigation. Perinatal care for women from ethnic minority backgrounds should focus on preventative measures to optimise maternal outcomes. TWEETABLE ABSTRACT: Women with an ethnic minority background giving birth in England have an increased risk of postpartum haemorrhage, even when characteristics of the mother, the baby and the care received are taken into account.