Saccharomyces cerevisiae is commonly used for industrial ethanol production. However, it cannot ferment xylose, the second most common fermentable sugar in the hydrolysates of lignocellulosic biomass. Meanwhile, xylulose, an isomerization product of xylose, can be metabolized by S. cerevisiae. S. cerevisiae transformed with the XYL1 and XYL2 genes encoding XR and XDH from Pichia stipitis and the XKS1 gene encoding xylulokinase (XK) from S. cerevisiae acquires the ability to ferment xylose to ethanol. However, this approach is insufficient for industrial bio-processes because of the low rate of fermentation and unfavorable excretion of xylitol. The difference in the coenzyme specificities of XR and XDH creates an intracellular redox imbalance that has been implicated as the main cause of xylitol excretion. To reduce xylitol formation during xylose fermentation, we have been examining laboratory recombinant S. cerevisiae strains that express a XDH mutant (ARSdR; D207A/I208R/F209S/N211R), which has a complete reversal of coenzyme specificity toward NADP⁺. In the present study, the flocculent industrial S. cerevisiae strain IR-2, which has high xylulose-fermenting ability, was first selected as a host suitable for genetically engineering xylose fermentation. We then constructed a recombinant strain (MA-R5) through the chromosomal integration of the NADP⁺-dependent XDH gene, as well as the XR and XK genes. MA-R5 had a markedly increased xylose consumption rate and an increased ethanol yield as compared to the reference strain (MA-R4) that expressed wild-type XDH. Furthermore, MA-R5 effectively co-fermented glucose and xylose and produced ethanol with a high yield from the detoxified hydrolysate of wood chips.