A recurring problem in watershed hydrology is the identification of flow paths along which water is routed in both surface and subsurface environments. In particular, the question of whether storm flow is composed primarily of water flowing over the land surface (overland flow) or beneath the land surface (subsurface flow) is significant, because the former flow path provides for little soil contact, while the latter allows for extensive interaction with subsurface materials. The Reedy Creek watershed (45.1 km2) in the Virginia Coastal Plain has been the subject of field hydrogeological, hydrochemical, and hydrological studies since 1989. Results from chemical separation of six storm flow hydrographs indicate that saturation overland flow (including both direct precipitation onto and return flow to the surface) from saturated contributing areas is the dominant storm flow generation mechanism in the watershed; new water composed 11--38% of creek discharge at the hydrograph peaks, and maximum 24--hour rainfall intensity explained 96% of the total variation in the peak new water contribution among the six events. Estimates of new water-contributing areas determined from the chemical separations were found to vary as a function of antecedent conditions and were also found to be consistent with estimates of areas of likely surface saturation (e.g., perennial channels, open water, and riparian wetland areas) based on field observations and topographic maps. ¿ American Geophysical Union 1993