Technical Abstract: The Bosque River Watershed in Texas is facing a suite of water quality issues including excess sediment, nutrient, and bacteria. The sources of the pollutants are improperly managed cropland and grazing land, dairy manure application, and effluent discharge from wastewater treatment facilities. Several best management practices (BMPs) have been proposed for pollution reduction and watershed protection. The overall objectives of this study were to demonstrate a modeling approach using Soil and Water Assessment Tool (SWAT) model to simulate various BMPs and assess their long-term impacts on sediment and nutrient loads at different spatial levels. The SWAT model was calibrated and validated for long-term annual and monthly flows at Valley Mills and for monthly sediment, total nitrogen (TN) and total phosphorus (TP) at Hico and Valley Mills monitoring locations. The BMPs including streambank stabilization, gully plugs, recharge structures, conservation tillage, terraces, contour farming, manure incorporation, filter strips, and PL-566 reservoirs were simulated in the watershed areas that met the respective practice’s specific criteria for implementation. These BMPs were represented in the pre- and post-conditions by modifying one or more channel parameters (channel cover, erodibility, Manning’s n), curve number (CN), support practice factor (P-factor), filter strip width, and tillage parameters (mixing efficiency, mixing depth). The BMPs were simulated individually and the resulting Hydrologic Response Units (HRUs), subwatershed, and watershed level impacts were quantified for each BMP. Sensitivity of model output values to input parameters used to represent the BMPs was also evaluated. Implementing individual BMPs reduced sediment loads from 3 percent to 37 percent and TN loads from 1 percent to 24 percent at the watershed outlet; however, the changes in TP loads ranged from 3 percent increase to 30 percent decrease. Higher reductions were simulated at the subwatershed and HRU levels. Among the parameters analyzed for sensitivity, P-factor and CN were most sensitive followed by Manning’s n. The TN and TP outputs were not sensitive to channel cover. This study showed that the SWAT modeling approach could be used to simulate and assess the effectiveness of agricultural best management practices.