Detection of water bodies and their flow paths is elemental process of terrain analysis in any application related with surface water. Several algorithms have been proposed for automatic extraction of surface flow paths from Digital Elevation Models (DEMs) (e.g., O’Callaghan and Mark, 1984; Tarboton, 1997; Paik, 2008). While these algorithms have been widely applied, DEMs-based approaches have fundamental limitation in detecting water bodies. To resolve this issue, we propose a hybrid approach that utilizes DEMs and satellite imagery together. These dataset supplement each other’s weakness, providing extra information for better detection of water bodies and flow paths. The proposed approach is tested for the Geum River, showing successfully extraction of flow boundaries and more resonable flow paths within water bodies.