The potential impact of climate change on fluvial flooding is receiving considerable scientific and political interest thanks to evidence from climate model projections and a widely held belief that flood risk may be increasing at European levels. This review compares published work on historical trends in UK rainfall and river flow records with high-resolution regional climate change projections, and attempts to reconcile apparent differences between the two. Attention is focused on the techniques used for climate change detection and attribution, as well as the potential confounding effects of land-use change. International and domestic efforts to build adaptive capacity rest on improved quantification of uncertainty in flood risk at very local, catchment and regional scales. This will involve further research to better integrate climate and land-management interactions, to understand changes in the dependence between different flood generating mechanisms, and to improve the characterization and communication of uncertainty at all stages of analysis. Resources are also needed to ensure that latest, but still uncertain, science is presented in an appropriate form to underpin policy development and is translated into sensible guidance for practitioners.