Anthropogenic activities increase levels of heavy metals in catchment soils while changing land use in a watershed and further impairing the receiving water. Using a watershed as a study unit might help to determine the contributions of various anthropogenic activities, or their driven land use changes to spatial loadings and environmental risks of heavy metals in catchment soils because of its close geographic boundary. Few studies have been performed at a watershed scale to understand spatial relationships between anthropogenic land use changes and heavy metal enrichment in catchment soils, and have combined this with an overall potential ecological risk of heavy metal enrichment. This study integrated satellite image interpretation, field surveys of soil heavy metal level, and geo-statistics to explore relationships of soil heavy metal loads and their potential ecological risk in an urbanizing watershed in Xiamen, China. The Bantou Reservoir watershed covers 205 km² and consists of forested land, agricultural land, urban land, green land, and water surface. Interpreted by the digital elevation map (DEM), the studied watershed is comprised of five subwatersheds. In total, 150 soil-sampling points were assigned in the watershed based on coverage of each land use type with a minimum of three samples in each subwatershed to satisfy statistical requirements. A composite sample of two top-10 cm soil samples was collected in a 10 ×10 m square at each sampling point. Six heavy metals, Cu, Zn, Pb, Cr, Ni, and Cd were analyzed in these topsoil samples. The enrichment factor and potential ecological risk factor of each heavy metal for each land use type, and total risk index of six heavy metals for each land use type were calculated. Results indicated that urban and agricultural land use significantly increased heavy metal loads in topsoils in comparison with forest and green land use in the watershed. The enrichment factor of heavy metals calculated on the basis of the regional background values showed that urban land use has higher values than agricultural land use, followed by green and forest land use. Accordingly, the potential ecological risk factor of each heavy metal followed the same pattern. Urban land use increased the enrichment of Cu, Zn, and Cd in topsoil of the watershed in comparison with other land uses. Lead enrichment factor was not related to land use type, suggesting that Pb was derived from atmospheric deposition. Regardless of land use types in the watershed, total Cr and Ni contents in topsoil were below the regional background values, although significant increases were observed within urban land in comparison with forest land. The land use based-inverse distance weighting (LU-IDW) spatial interpolation revealed that heavy metal loads and potential ecological risks are clearly related to anthropogenic activities, especially to urbanization. This study demonstrated that the methodology we applied to understand relationships between land use change with urbanization and heavy metal enrichment and even their ecological risks at a watershed scale is feasible. The integration of traditional field survey and satellite image interpretation techniques can serve as a powerful methodology for watershed-scale studies on environmental changes.