Greenhouse gas (GHG) emissions throughout the biofuels life cycle stem from many sources, such as feedstock. cultivation, transportation, production, and land use change (LUC). No work has identified biomass sourcing and ethanol productions strategies with state-of-the-art LUC models to minimise life cycle GHG emissions of biofuels including emissions from LUC. We integrate state-of-the-art land use modelling methodology with mathematical programming techniques to identify optimal global biomass sourcing, transportation, and ethanol production strategies that minimise life cycle GHG emissions. We consider a case study where the US demands 189 GL/y of ethanol. Switchgrass is the most chosen feedstock by far (89%), followed by corn stover (8%) and a small amount of corn (3%). The US produces most of the required ethanol (156 GL/y) with switchgrass grown on marginal land in the central and western US. China (16.3 GL/y), Brazil (8.18 GL/y), the EU (5.22 GL/y), Central America (3.10 GL/y), and India (0.64 GL/y) also produce ethanol and ship it to the US. Total life cycle emissions are 43.2 g CO2-eq/MJ, and LUC emissions account for only 1.1 g CO2-eq/M.1. Most life cycle emissions stem from direct processing (32.8 g CO2-eq/MJ).