AnnAGNPS was applied to the Cedar Creek Watershed and its Matson Ditch sub-catchment in the St. Joseph River Watershed to assess its effectiveness at predicting monthly hydrology and pesticide loading at different scales. The model was also utilized to assess the effects that conservation tillage has on monthly stream flow and pesticide loading at both watershed and sub-catchment scales. Calibration of the AnnAGNPS model for monthly hydrology of the Cedar Creek Watershed between January 1989 and December 1998 was accomplished by adjusting runoff curve numbers. Model validation compared observed stream flow from the years 1999-2006 to the AnnAGNPS derived predications for the respective years. Land use was predominately agricultural with the majority of the land being in a corn to soybean crop rotation. The model utilizes a 1/3 arc second United States Geological Survey digital elevation model, re-sampled to exactly 10 meter grid size, to determine the topographical and hydrologic characteristics in each watershed. Soil parameters were constructed from the SSURGO soil data while land-use and cover was obtained from National Agricultural Statistics Service data for 2001. Once the land use was determined, representative management schedules were created by incorporating county data for the different management practices and their respective dates. A five-year management plan constructed from the NASS and county data was utilized in two cycles to create a 10-year simulation. In addition, a two-year initialization period was implemented to synchronize the management rotations. Weather inputs were constructed using data from the National Climactic Data Center weather station located in Garret, Indiana. This study improves our understanding of the effectiveness of calibrating the AnnAGNPS model and its effects on the prediction capabilities of the model when applied at different scales. Additionally, it increased our understanding of conservation tillage effects at multiple scales.