Growing biofuel feedstocks on prime cropland creates pressure to both intensify and expand agriculture elsewhere via changes in the prices of commodities and land. This tends to increase greenhouse gas (GHG) emissions, but this market-mediated land use change effect is neglected in life-cycle assessments of biofuels. The 2007 Energy Bill may change this situation because it sets ambitious GHG emission reduction goals for advanced biofuels and explicitly requires including these indirect effects. Preliminary estimates indicate that the loss of carbon stocks from off-site land conversion is a large contributor to greenhouse emissions associated with crop-based biofuels, potentially outweighing all other emissions sources combined. Thus, a better understanding of land use change associated with biofuels will greatly inform the development of biofuel technologies and policies that will be compatible with climate change mitigation. The economic and bio-physical models needed to estimate market-mediated land use change effects are fraught with methodological challenges and data uncertainties. We review existing models of land use change and current estimates of indirect GHG emissions in order to quantify the challenge for crop-based biofuels, and to identify the methodological differences and data uncertainties they contain. We also present a research strategy for reducing the uncertainties in these estimates and developing more usable modeling tools.