This paper describes a geographic information system (GIS) approach to modeling the effects of distance to water and slope angle on external phosphorus loading to lakes. A raster GIS database was created that included land use, topography, soils, and watershed boundaries for three Indiana reservoirs (Lakes Kickapoo, Lenape, and Shakamak). Three sets of phosphorus export coefficients were selected and assigned to each cell according to land use. Linear filters were designed and applied to the export coefficients such that areas nearest the water and with the steepest slopes would contribute the greatest amount of phosphorus relative to the initial phosphorus export coefficient. These filtered coefficients were used to calculate aerial phosphorus loading and in-lake phosphorus concentrations. The predicted concentrations were within 5 μg/L of the observed phosphorus concentration in Lake Shakamak for three modeling scenarios, and within 22 μg/L of the observed concentration in Lake Lenape for three scenarios. Predictions of phosphorus concentrations in Lake Kickapoo were consistently low (35 percent to 95 percent). This may have resulted from the complex hydrology of Lake Kickapoo or the inability to accurately model the physical processes as intended. The GIS system was useful for modeling the effects of distance and slope on phosphorus loading and for providing data highlighting critical management areas.