To probe the catalytic mechanism of human purine nucleoside phosphorylase (PNP), 13 active-site mutants were constructed and characterized by steady-state kinetics. In addition, microtiter plate assays were developed for both the phosphorolytic and synthetic reactions and used to determine the kinetic parameters of each mutant. Mutations in the purine binding site exhibited the largest effects on enzymatic activity with the Asn243Ala mutant resulting in a 1000-fold decrease in the kcat for inosine phosphorolysis. This result in combination with the crystallographic location of the Asn243 side chain suggested a potential transition state (TS) structure involving hydrogen bond donation by the carboxamido group of Asn243 to N7 of the purine base. Analogous to the oxyanion hole of serine proteases, this hydrogen bond was predicted to aid catalysis by preferentially stabilizing the TS as a consequence of the increase in negative charge on N7 that occurs during glycosidic bond cleavage and the associated increase in the N7-Asn243 hydrogen bond strength. Two residues in the phosphate binding site, namely His86 and Glu89, were also predicted to be catalytically important based on their alignment with phosphate in the X-ray structure and the 10-25-fold reduction in catalytic activity for the His86Ala and Glu89Ala mutants. In contrast, catalytic efficiencies for the Tyr88Phe and Lys244Ala mutants were comparable with wild-type, indicating that the hydrogen bonds predicted in the initial X-ray structure of PNP [Ealick, S. E., et al. (1990) J. Biol. Chem. 265, 1812-1820] were not essential for catalysis. These results provided the foundation for studies reported in the ensuing two manuscripts focused on the PNP catalytic mechanism [Erion, M. D., et al. (1997) Biochemistry 36, 11735-11748] and the use of mutagenesis to reverse the PNP substrate specificity from 6-oxopurines to 6-aminopurines [Stoeckler, J. D., et al. (1997) Biochemistry 36, 11749-11756].