Interest in bioenergy crops is increasing due to their potential to reduce greenhouse gas emissions and dependence on fossil fuels. We combined process-based and geospatial models to estimate the potential biomass productivity of miscanthus and its potential impact on soil carbon stocks in the croplands of the continental United States. The optimum (climatic potential) rainfed productivity for field-dried miscanthus biomass ranged from 1 to 23Mgbiomassha-1yr-1, with a spatial average of 13Mgha-1yr-1 and a coefficient of variation of 30%. This variation resulted primarily from the spatial heterogeneity of effective rainfall, growing degree days, temperature, and solar radiation interception. Cultivating miscanthus would result in a soil organic carbon (SOC) sequestration at the rate of 0.16-0.82MgCha-1yr-1 across the croplands due to cessation of tillage and increased biomass carbon input into the soil system. We identified about 81millionha of cropland, primarily in the eastern United States, that could sustain economically viable (>10Mgha-1yr-1) production without supplemental irrigation, of which about 14millionha would reach optimal miscanthus growth. To meet targets of the US Energy Independence and Security Act of 2007 using miscanthus as feedstock, 19millionha of cropland would be needed (spatial average 13Mgha-1yr-1) or about 16% less than is currently dedicated to US corn-based ethanol production.