Abstract: Currently, the predominant microbially produced biofuel is starch- or sugar-derived ethanol. However, ethanol is not an ideal fuel molecule, and lignocellulosic feedstocks are considerably more abundant than both starch and sugar. Thus, many improvements in both the feedstock and the fuel have been proposed. In this paper, we examine the prospects for bioproduction of four second-generation biofuels (n-butanol, 2-butanol, terpenoids, or higher lipids) from four feedstocks (sugars and starches, lignocellulosics, syngas, and atmospheric carbon dioxide). The principal obstacle to commercial production of these fuels is that microbial catalysts of robust yields, productivities, and titers have yet to be developed. Suitable microbial hosts for biofuel production must tolerate process stresses such as end-product toxicity and tolerance to fermentation inhibitors in order to achieve high yields and titers. We tested seven fast-growing host organisms for tolerance to production stresses, and discuss several metabolic engineering strategies for the improvement of biofuels production. [Copyright &y& Elsevier]