Horseradish peroxidase (HRP) with a variety of potential biotechnological applications is still isolated from the horseradish root as a mixture of different isoenzymes with different biochemical properties. There is an increasing demand for preparations of high amounts of pure enzyme but its recombinant production is limited because of the lack of glycosylation in Escherichia coli and different glycosylation patterns in yeasts which affects its stability parameters. The goal of this study was to increase the stability of non-glycosylated enzyme, which is produced in E. coli, toward hydrogen peroxide via mutagenesis. Asparagine 268, one of the N-glycosylation sites of the enzyme, has been mutated via saturation mutagenesis using the megaprimer method. Modification and miniaturization of previously described protocols enabled screening of a library propagated in E. coli XJb (DE3). The library of mutants was screened for stability toward hydrogen peroxide with azinobis (ethylbenzthiazoline sulfonate) as a reducing substrate. Asn268Gly mutant, the top variant from the screening, exhibited 18-fold increased stability toward hydrogen peroxide and twice improved thermal stability compared with the recombinant HRP. Moreover, the substitution led to 2.5-fold improvement in the catalytic efficiency with phenol/4-aminoantipyrine. Constructed mutant represents a stable biocatalyst, which may find use in medical diagnostics, biosensing, and bioprocesses.