In this paper, we discuss the importance of developing integrated assessment models to support the design and implementation of policies to address water quality problems associated with agricultural pollution. We describe a new modeling system, LUMINATE, which links land use decisions made at the field scale in the Upper Mississippi, Ohio, and Tennessee basins through both environmental and hydrologic components to downstream water quality effects and hypoxia in the Gulf of Mexico. This modeling system can be used to analyze detailed policy scenarios identifying the costs of the policies and their resulting benefits for improved local and regional water quality. We demonstrate the model’s capabilities with a simple scenario where cover crops are incentivized with green payments over a large expanse of the watershed.