Each state and district within the Chesapeake Bay watershed has cooperated with the Chesapeake Bay Program (CBP) to develop local Watershed Implementation Plans (WIPs) that identify the type and quantity of best management practices (BMPs) that, if implemented, are estimated to meet 2025 Total Maximum Daily Load (TMDL) goals for Bay water quality. However, top-down management of large regions, such as the 167,000-km2 Bay catchment, is often necessarily limited by the feasibility of providing implementation plans that are customized by watershed hydro-physiographic characteristics and socio-political considerations. The Bay simulation model divides the catchment into watersheds of approximately 350 km2 each; these watersheds become the Bay model’s smallest overland management unit. We used Bay WIP plans, local information, and a hydrologic model called Topo-SWAT to model three of these smallest-unit watersheds in more local detail. Our smallest management unit became contiguous, similarly managed, cropland areas (i.e., one or several neighboring agricultural fields) and these management units were further divided by the topographic wetness index. Our watersheds represent three distinct hydrological and geochemical regions within the Chesapeake Bay catchment, namely Appalachian Valley and Ridge – karst, Appalachian Valley and Ridge – nonkarst, and Appalachian Piedmont. We modeled three scenarios for each watershed: baseline (pre-WIP), WIP implementation, and “smarter” WIP placement where we targeted BMP placements for cost-effectiveness. We then compared results among scenarios as well as across watersheds. We are interested to see how well the models agree at the watershed outlet, discover cost-effective BMP placements within each watershed that meet WIP goals, and compare our findings across the physiographic regions to determine how they can guide regional planning.