Life cycle assessment (LCA) has shown that first generation biofuels provide a little to no benefit for greenhouse gas (GHG) reductions compared to petroleum fuels, particularly when indirect effects are considered. Second generation fuels are intended to achieve greater GHG reductions and avoid other sustainability issues. LCAs of second generation biofuels exhibit great variability and uncertainty, leading to inconclusive results for the performance of particular pathways (combinations of feedstocks and fuels). Variability arises in part because of the prospective nature of LCAs for future fuels; however, a review of recent articles on biofuel LCA methodology indicates two additional sources of variability: real sources such as spatiotemporal heterogeneity, and methodological sources such as choices for co-product allocation methods and system boundary definition.