Recent model developments enable the use of higher resolution representation of certain sectors and geographic regions in Integrated Assessment Models (IAMs) with the advantage of maintaining the global scope of analysis and resultant system interactions. Such information can then be downscaled to even higher resolutions for use in further scientific analysis or decision support applications. However, higher resolution nested simulations involve a cost of data assembly and reconciliation as well as new model calibration. Here we test the importance of spatial resolution of the Global Change Assessment Model (GCAM) agriculture and land use sector for the US Midwest. The standard GCAM spatial resolution makes use of 10 land use regions in the USA, several of which overlap with the 14 state region of interest in this study. In previous work we disaggregated the study region into 37 land use regions based on political (state) and agricultural (crop management zone, CMZ) boundaries. For the simulations presented here, we use a common set of data inputs and aggregate these to 1, 12 (CMZ), 14 (state) and 37 (combination of states and CMZs) regions, respectively. Our objective is to evaluate and compare the model's future agricultural sector results in these different spatial configurations, in reference (no climate policy) scenarios and in response to climate mitigation policies. The mitigation policies assessed here put an economic price on all CO2 emissions from fossil fuels and from land use change such that end-of-century radiative forcing is stabilized at 4.5 W/m2. We find that the spatial aggregation does not alter the basic relationship between the reference and mitigation policy pairs. However the different spatial scale does result in differences in the relative future extent of food and bioenergy crops within the study region. We plan to extend this analysis further by downscaling future land cover to a high resolution grid and evaluating the importance of the different regional aggregation of the modeled land use regions for the final, spatially explicit, model product. This analysis will help inform selection of appropriate spatial scale of analysis for agriculture and land use studies in integrated assessment models.