ObjectiveTo assess whether secondary prevention, which preemptively treats women with above‐average postpartum bleeding, is non‐inferior to universal prophylaxis.DesignA cluster‐randomised non‐inferiority community trial.SettingHealth sub‐centres and home deliveries in the Bijapur district of Karnataka, India.PopulationWomen with low‐risk pregnancies who were eligible for delivery with an Auxiliary Nurse Midwife at home or sub‐centre and who consented to be part of the study.MethodsAuxiliary Nurse Midwifes were randomised to secondary prevention using 800 mcg sublingual misoprostol administered to women with postpartum blood loss ≥350 ml or to universal prophylaxis using 600 mcg oral misoprostol administered to all women during the third stage of labour.Main outcome measuresPostpartum haemoglobin ≤7.8 g/dl, mean postpartum blood loss and postpartum haemoglobin, postpartum haemorrhage rate, transfer to higher‐level facilities, acceptability and feasibility of the intervention.ResultsMisoprostol was administered to 99.7% of women as primary prevention. In secondary prevention, 92 (4.7%) women had postpartum bleeding ≥350 ml, of which 90 (97.8%) received misoprostol. The proportion of women with postpartum haemoglobin ≤7.8 g/dl was 5.9 and 8.8% in secondary and primary prevention clusters, respectively [difference −2.9%, one‐sided 95% confidence interval (CI) <1.3%]. Postpartum transfer and haemorrhage rates were low (<1%) in both groups. Shivering was more common in primary prevention clusters (P = 0.013).ConclusionSecondary prevention of postpartum haemorrhage with misoprostol is non‐inferior to universal prophylaxis based on the primary outcome of postpartum haemoglobin. Secondary prevention could be a good alternative to universal prophylaxis as it medicates fewer women and is an acceptable and feasible strategy at the community level.Tweetable abstractSecondary prevention of postpartum haemorrhage with misoprostol is non‐inferior to universal prophylaxis.