A conventional, photogrammetrically-derived digital elevation model (DEM) (10 m resolution) and a Light Detection and Ranging (LiDAR)-derived DEM (1 m resolution) were used to model the stream network and soil moisture conditions of a 193 ha watershed in the Swan Hills of Alberta, Canada. Soil moisture was modeled spatially using an innovative GIS algorithm based on topography. The time-averaged tendency for soil to be saturated is indicated by a depth-to-water value assigned on the basis of distance to a surface water feature and the slope of the land surface. The actual stream network and distribution of wet soils, mapped in the field, were used as verification. Modeled stream networks and soil moisture conditions were found to be generally consistent with the field-mapped wet soil areas. The LiDAR DEM-modeled stream network and soil moisture conditions was the most accurate representation of the field-mapped situation. A kmatch criteria showed that 78 % of the field-mapped wet area was fully embedded in the 0 to 2 m depth-to-water class modeled using the LiDAR DEM. Using the conventional DEM, 56 % of the wet area was within the 0 to 3 m depth-to-water class. All major areas of wet soils and their hydrologic connectivity were successfully modeled using the LiDAR DEM, whereas those modeled using the conventional DEM missed some. This was likely due to the greater initial point density, accuracy and resolution of the LiDAR DEM compared to the conventional DEM. The DEM-based algorithm detailed here has great potential for application in areas such as distributed hydrologic modelling, forestry and other areas of natural resource management, and modelling of development impacts on wetlands and riparian zones.