Glycine oxidase (GOX) from Bacillus subtilis is a new flavoprotein of great potential biotechnological use that catalizes the oxidative deamination of various amines (glycine, sarcosine, and N‐ethyl‐glycine) and d‐amino acids (d‐alanine and d‐proline). However, its commercial application is hindered by its low heterologous expression in Escherichia coli due to its codon bias and the sensitivity of its N‐terminus to proteases. The first problem has been solved by cloning the GOX gene from B. subtilis ATCC 6633 into the Rosetta E. coli strain, which contains the pRARE plasmid. The second problem was overcome by inserting the gene in the pET28a expression vector, which not only has a 6×His tag but also increases the N‐terminus in 36 amino acids without impairing either the enzymatic activity or the ribosome binding region. After induction with 0.5 mM isopropyl thio‐β‐d‐galactoside for 5 h in TB‐medium, the soluble and active chimeric GOX was expressed up to 15.81 U·g−1 cell, with a fermentation yield of 399 U·L−1. The latter value represents about 16% of the total soluble protein content of the cell. The three latter values are higher than the best found in the literature by 16‐, 28‐ and 4‐fold, respectively. The enzyme was purified with a nickel HiTrap chelating‐affinity column in 96% yield to apparent homogeneity. It was fully active and was stable for months at –80 °C in the presence of 10% glycerol. Its substrate specificity was similar to that previously described, but the constructed M261 mutants unexpectedly decreased in KM compared with the wild‐type, especially in the M261Y mutant. Noteworthy, there was decrease in the KM for N‐ethyl‐glycine of up to 0.7 mM, similar to that found with N‐alkyl‐glycine oxidase. Such mutants open up new possible uses of this enzyme not only in the pharmacological industry but also in the clinical field for diabetic complications.