In recent years there has been a flurry of activity aimed at evaluating the land use consequences of biofuels programs and the associated carbon releases. In this paper we argue that these studies have tended to underestimate the ensuing land use emissions, because they have ignored the role of irrigation, and associated constraints on cropland expansion. In this paper, we develop a new general equilibrium model which distinguishes irrigated and rain fed cropping industries at a global scale. Using the new model we evaluate the implications of land use change due to US ethanol programs, in the context of physical constraints on the expansion of irrigated cropland. We find that models which mingle irrigated and rain fed areas underestimate the global land use changes induced due to the US ethanol expansion by about 5.7%. They tend to underestimate the corresponding land use emissions by more than one fifth.