AGNPS, a nonpoint source pollution model, was selected to stimulate sediment yield and chemical loadings from Owl Run watershed. The model was validated to demonstrate its applicability to Virginia Piedmont conditions. The validation was carried out by comparing simulation results with measured data including runoff, sediment yield, and nitrogen and phosphorus loadings to downstream water bodies. Statistical measures, including simple linear regression, determination of root mean square errors, and test on differences between simulated and measured data, were used in this study to evaluate errors. Results from these statistical procedures indicated that the errors between simulated and measured results are within acceptable limits. An annualization procedure was used to provide the basis for evaluating the long-term impact of various BMP’s. Critical areas in the watershed, which are responsible for majority of the pollutant loadings, were identified by the model using the annulization procedure. A FORTRAN program was developed to convery critical areas for individual events to “annualized critical areas” so that evaluations were made on long-term basis. BMPs currently installed in Owl Run watershed and several alternative BMP implementation scenarios were simulated. Their impacts on reducing pollutant loadings and their cost effectiveness were evaluated by using the AGNPS model and the annualization procedure. The current BMP scenario will eventually reduce sediment yield, total nitrogen, and total phosphorus loadings by 26%, 32%, and 32% respectively. Some of the proposed scenarios can reduce these pollutant loadings by up to 59%, 66%, and 67% respectively. / Master of Science