Community resilience is increasingly central to climate adaptation policy, yet household-level evidence from riverine island (char) settings remains limited. This study assessed household flood resilience in riverine island communities of northern Bangladesh and developed a context-specific Flood Resilience Index (FRI) using targeted literature review, expert consultation, and repeated char field engagement (pilot interviews and village focus group discussions). We conducted a cross-sectional household survey (n = 384) in six villages across Ulipur Upazila and Chilmari Upazila, Kurigram District, Bangladesh, during July–September 2023 to populate the index. The FRI was constructed from 48 indicators grouped into eight components and three capacities (anticipatory, absorptive, adaptive) using min–max normalization and equal (unit) weights; capacity and overall scores were obtained via arithmetic aggregation across the 3A structure, and radar/spider plots were used to visualize and compare capacity/component profiles across upazilas. The overall flood resilience index was 0.326, indicating that char dwellers are less resilient against recurrent floods irrespective of where they live. The anticipatory, absorptive, and adaptive capacity indices were 0.333, 0.289, and 0.339, indicating generally low flood resilience; households closer to the mainland were more resilient than those farther away, reflecting better communication, services, and livelihood opportunities. Policy priorities include improving access to food and healthcare, income diversification, actionable early warning and risk information, and resilient connectivity/infrastructure through coordinated local government action, NGO participation, and public–private partnerships.