Flooding is one of the most costly natural hazards, inflicting billions of dollars in annual damages. Low-lying coastal regions are particularly prone to flood risks, where floods may arise from a combination of rising sea levels, storm surges, high tides, river flooding (fluvial), heavy rainfall (pluvial), and land surface changes (e.g., increased impervious surface area). The interaction of these flood drivers leads to flood dynamics that can vary greatly over short distances and timescales, making it challenging to accurately predict coastal urban flood risks. Supported by the Department of Energy’s Integrated Coastal Modeling Project (ICoM), we have developed a high-resolution, high-fidelity, regional-to-urban flood hazard modeling framework. This framework brings together regional hydrology (DHSVM), coastal hydrodynamics (FVCOM), and urban hydrodynamics (RIFT) models to represent the complex interactions among riverine, coastal, and urban flood processes. Using the DHSVM-FVCOM-RIFT framework, we assessed region-to-urban flood hazards and their drivers in two distinct U.S. coastal regions and their metropolitical centers: multi-decadal flooding events in the Delaware-Philadelphia area and a compound flooding event in Puget Sound near Seattle. These models, driven by kilometer- or near-kilometer-scale climate simulations, consistently demonstrate strong predictive capabilities in river, coastal, and urban environments, validated against gauge observational datasets. Notably, the framework uniquely provides property-level assessments of urban flood hazards and identifies event-specific flood drivers (e.g., surge-only, pluvial-fluvial, fluvial-surge-pluvial), offering actionable insights for developing targeted, effective flood mitigation strategies.