A spatially distributed hydrological model, based on the MIKE-SHE code, has been constructed for the 375 000 km2 Senegal River basin in West Africa (Andersen et al., 2001). Calibration and validation results suggest that the spatial resolution of the input data in parts of the area has not been sufficient for satisfactory modelling performance. In order to improve the model results, especially the spatial output, a remotely sensed dryness index (Sandholt et al., 2001) has been visually examined for its relation to model soil moisture for two consecutive days in the wet season of 1990. The visual comparison showed the expected consistency in both root zone soil moisture and dryness index between the two days. An examination of the spatial patterns in model input and output variables revealed the input of precipitation zones to be highly dominating on the pattern of the output variables.