Despite the widespread use of GIS over the past ten years, it has had limited application for statewide modeling of pollutant loadings such as sediment, nitrogen, and phosphorus. Related literature on possible approaches to statewide modeling is also limited. In this study a GIS-based screening model was used to rank 104 watersheds in Pennsylvania for nonpoint pollution potential. The goals of this project were to: select important processes and parameters of watersheds that contribute to water quality degradation, develop appropriate statewide data, develop a ranking model to use the data and verify the model by comparing results with existing water quality data for specific watersheds. The GIS database consisted of watershed boundaries, land cover, animal density, topography, soils, precipitation and rainfall erosivity. The 104 watersheds were ranked by the Agricultural Pollution Potential Index (APPI) - a ranking model comprised of the following four separate components: the Sediment Production Index (SPI) assesses the potential interrill and rill erosion and sediment delivery from field to stream, the Runoff Index (RI) predicts the watershed's potential to produce surface runoff, the Animal Loading Index (ALI) ranks the potential manure production within a watershed, the Chemical Use Index (CUI) ranks the potential for agricultural pollution production from use of commercial chemical applications to the various agricultural lands. Ranking watersheds in Pennsylvania by their agricultural pollution potential was completed and critical pollution production areas were identified.