The growth in federal conservation programs has created a need for policy modeling frameworks capable of measuring micro-level behavioral responses and macro-level landscape changes. This paper presents an empirical model that predicts crop choices, crop rotations, and conservation tillage adoption as a function of conservation payment levels, profits, and other variables at more than 42,000 agricultural sites of the National Resource Inventory (NRI) in the Upper Mississippi River Basin. Predicted changes in crop choices and tillage practices are then fed into site-specific environmental production functions to determine changes in nitrate runoff and leaching and in water and wind erosion at each NRI site. This policy-scale model is applied to the case of green payments for the adoption of conservation practices (conservation tillage and crop rotations) in the Upper Mississippi River Basin, a region under scrutiny as a significant source of nutrient loadings to the Mississippi River, causing hypoxia in the Gulf of Mexico. Results from this application suggest that payments for conservation tillage and crop rotations increase the use of these conservation practices. However, the acreage response is inelastic and the programs are not likely to be cost effective on their own for addressing the hypoxia problem in the Gulf of Mexico.