Flood risk is increasingly shaped by intensifying rainfall extremes interacting with complex catchment processes under a changing climate. This chapter brings together advances in modelling frameworks that link extreme rainfall to flood generation and risk assessment. It reviews recent developments in hydrological and hydraulic modelling, including machine learning and physics-informed approaches that improve prediction across diverse catchments, alongside efficient two-dimensional hydrodynamic models that enable large-scale flood mapping. The chapter also discusses methods such as statistical downscaling and weather generators that translate climate-scale information into local flood impacts. Despite these advances, challenges remain in representing non-stationarity, parameter uncertainty, and model structural limitations under evolving climate conditions. Together, these modelling approaches improve our ability to simulate flood risk across scales, while also highlighting key uncertainties that must be addressed to support robust climate adaptation and risk-informed decision-making.