Lignocellulosic biomass waste is an abundant renewable resource of sugars for fermentation to biofuel. Due to its high fermentation capability and tolerance to ethanol and inhibitors, Saccharomyces cerevisiae was chosen to engineer a strain able to ferment xylose to ethanol. Wild-type S. cerevisiae is not able to grow on xylose as a sole carbon source. In xylose-fermenting yeasts, xylose reductase reduces the sugar to xylitol, which is then oxidized to xylulose by a xylitol dehydrogenase. These two enzymes require different cofactors, which leads to a cofactor imbalance in wild-type cells attempting to utilize xylose. In order to bypass the two-step oxidoreductive isomerization reaction, the xylose reductase-encoding GRE3 gene was knocked out and the pathway was replaced with a xylose isomerase (XYLA) isolated from Piromyces sp. E2, to convert xylose directly to xylulose. To increase the flux of xylulose towards the pentose phosphate pathway, a second copy of the endogenous xylulokinase (XKS1) was constitutively expressed. This two-gene construct was chromosomally integrated into the GRE3 deletion strains and the resulting strain was able to grow aerobically on xylose as its sole carbon source. Anaerobic glucose-xylose co-fermentation experiments yielded increased growth as compared to glucose only cultures, but ethanol production did not increase. In micro-aerobic high cell density fermentation the strains successfully produced ethanol from xylose as its sole carbon source and from a mixture of glucose and xylose. Evolutionary engineering further improved the growth rate, ethanol yield and specific ethanol productivity of these strains.