SummaryBackgroundAround 105 million people in India will be living in informal settlements by 2017. We investigated the effects of local resource centres delivering integrated activities to improve women's and children's health in urban informal settlements.MethodsIn a cluster-randomised controlled trial in 40 clusters, each containing around 600 households, 20 were randomly allocated to have a resource centre (intervention group) and 20 no centre (control group). Community organisers in the intervention centres addressed maternal and neonatal health, child health and nutrition, reproductive health, and prevention of violence against women and children through home visits, group meetings, day care, community events, service provision, and liaison. The primary endpoints were met need for family planning in women aged 15–49 years, proportion of children aged 12–23 months fully immunised, and proportion of children younger than 5 years with anthropometric wasting. Census interviews with women aged 15–49 years were done before and 2 years after the intervention was implemented. The primary intention-to-treat analysis compared cluster allocation groups after the intervention. We also analysed the per-protocol population (all women with data from both censuses) and assessed cluster-level changes. This study is registered with ISRCTN, number ISRCTN56183183, and Clinical Trials Registry of India, number CTRI/2012/09/003004.Findings12 614 households were allocated to the intervention and 12 239 to control. Postintervention data were available for 8271 women and 5371 children younger than 5 years in the intervention group, and 7965 women and 5180 children in the control group. Met need for family planning was greater in the intervention clusters than in the control clusters (odds ratio [OR] 1·31, 95% CI 1·11–1·53). The proportions of fully immunised children were similar in the intervention and control groups in the intention-to-treat analysis (OR 1·30, 95% CI 0·84–2·01), but were greater in the intervention group when assessed per protocol (1·73, 1·05–2·86). Childhood wasting did not differ between groups (OR 0·92, 95% CI 0·75–1·12), although improvement was seen at the cluster level in the intervention group (p=0·020).InterpretationThis community resource model seems feasible and replicable and may be protocolised for expansion.FundingWellcome Trust, CRY, Cipla.