Abstract Background Health needs segmentation is increasingly used in healthcare delivery and health policy planning to cater for diverse population groups and develop effective and equitable health policies. But existing models often lack a systemic approach, focusing instead on single conditions or settings. This leads to fragmentation, gaps and inefficiencies in service provision and inequities in health outcomes, particularly when addressing complexity, and specifically for children, as adversity clusters in families. Aims To develop a household health needs segmentation model to identify those households with children with high levels of complex cross-sectoral needs, to plan for integrated care services. Methods Data linkage of primary, secondary, mental, community and social care services, mortality records, residential information, and small-area multiple deprivation measures was used to design a rule-based model to identifying households with children with the most complex needs in the UK region of Cheshire and Merseyside (C&M), with large variation in demography, (human) geography, and deprivation. Results Of 2,645,329 individuals in C&M (97% of the registered population in 2021), 1,022,840 lived in 266,939 households with children aged 0-16: 21,527 households had complex needs. This 8% of families accounted for an estimated 34% of health and social care costs for families, £362 million in total, 42% of which spent on children in care of local authorities. Implications The model identified a specific group of families with complex service use patterns, high mental and physical comorbidity and socioeconomic vulnerabilities. These exhibit potential for improved outcomes through better integration of services and targeting of community building and family support resources. Conclusions Our model represents a significant innovation in health needs segmentation. By using whole system linked data, it provides a pragmatic way of profiling complexity for proactive care. Key messages • Complex health and social problems in children cluster in families with adults also having complex needs, so we need to transform fragmented services to support the whole family to improve outcomes. • Household level segmentation using cross-sectoral, whole system data is an innovation to identify households with complex needs with respect to their locality to improve integration of support.