The strict cofactor specificity of many enzymes can potentially become a liability when these enzymes are to be employed in an artificial metabolic pathway. The preference for NADPH over NADH exhibited by the Corynebacterium 2,5-diketo-D-gluconic acid (2,5-DKG) reductase may not be ideal for use in industrial scale vitamin C biosynthesis. We have previously reported making a number of site-directed mutations at five sites located in the cofactor-binding pocket that interact with the 2'-phosphate group of NADPH. These mutations conferred greater activity with NADH upon the Corynebacterium 2,5-DKG reductase [Banta, S., Swanson, B. A., Wu, S., Jarnagin, A., and Anderson, S. (2002) Protein Eng. 15, 131-140; (1)]. The best of these mutations have now been combined to see if further improvements can be obtained. In addition, several chimeric mutants have been produced that contain the same residues as are found in other members of the aldo-keto reductase superfamily that are naturally able to use NADH as a cofactor. The most active mutants obtained in this work were also combined with a previously reported substrate-binding pocket double mutant, F22Y/A272G. Mutant activity was assayed using activity-stained native polyacrylamide gels. Superior mutants were purified and subjected to a simplified kinetic analysis. The simplified kinetic analysis was extended for the most active mutants in order to obtain the kinetic parameters in the full-ordered bi bi rate equation in the absence of products, with both NADH and NADPH as cofactors. The best mutant 2,5-DKG reductase produced in this work was the F22Y/K232G/R238H/A272G quadruple mutant, which exhibits activity with NADH that is more than 2 orders of magnitude higher than that of the wild-type enzyme, and it retains a high level of activity with NADPH. This new 2,5-DKG reductase may be a valuable new catalyst for use in vitamin C biosynthesis.