OBJECTIVE To evaluate the effectiveness of the systematic use of a transparent plastic collector bag to measure postpartum blood loss after vaginal delivery in reducing the incidence of severe postpartum haemorrhage. DESIGN Cluster randomised trial. SETTING 13 European countries. PARTICIPANTS 78 maternity units and 25 381 women who had a vaginal delivery. INTERVENTIONS Maternity units were randomly assigned to systematic use of a collector bag (intervention group) or to continue to visually assess postpartum blood loss after vaginal delivery (control group). MAIN OUTCOME MEASURES The primary outcome was the incidence of severe postpartum haemorrhage in vaginal deliveries, defined as a composite of one or more of blood transfusion, intravenous plasma expansion, arterial embolisation, surgical procedure, admission to an intensive care unit, treatment with recombinant factor VII, and death. RESULTS Severe postpartum haemorrhage occurred in 189 of 11 037 of vaginal deliveries (1.71%) in the intervention group compared with 295 of 14 344 in the control group (2.06%). The difference was not statistically significant either in individual level analysis (adjusted odds ratio 0.82, 95% confidence interval 0.26 to 2.53) or in cluster level analysis (difference in weighted mean rate adjusted for baseline rate 0.16%, 95% confidence interval -0.69% to 1.02%). CONCLUSION Compared with visual estimation of postpartum blood loss the use of a collector bag after vaginal delivery did not reduce the rate of severe postpartum haemorrhage. TRIAL REGISTRATION Current Controlled Trials ISRCTN66197422.