Arginine deiminase (ADI) is a therapeutic protein for cancer therapy of arginine-auxotrophic tumors. However, its application as anticancer drug is hampered by its poor stability under physiological conditions in the bloodstream. Commonly, random PEGylation is being used for increasing the stability of ADI and in turn the improved half-life. However, the traditional random PEGylation usually leads to poor PEGylation efficiency due to the limited number of Lys on the protein surface. To boost the PEGylation efficiency and enhance the stability of ADI further, surface engineering of PpADI (an ADI from Pseudomonas plecoglossicida) to increase the suitable PEGylation sites was carried out. A new in silico approach for increasing the PEGylation sites was developed. The validation of this approach was performed on previously identified PpADI variant M31 to increase potential PEGylation sites. Four Arg residues on the surface of PpADI M31 were selected through three criteria and subsequently substituted to Lys, aiming for providing primary amines for PEGylation. Two out of the four substitutions (R299K and R382K) enhanced the stability of PEGylated PpADI in human serum. The average numbers of PEGylation sites were increased from ~12 (tetrameric PpADI M31, starting point) to ~20 (tetrameric PpADI M36, final variant). Importantly, the PEGylated PpADI M36 after PEGylation exhibited significantly improved T-m values (M31: 40 degrees C; M36: 40 degrees C; polyethylene glycol [PEG]-M31: 54 degrees C; PEG-M36: 64 degrees C) and half-life in human serum (M31: 1.9 days; M36: 2.0 days; PEG-M31: 3.2 days; PEG-M36: 4.8 days). These proved that surface engineering is an effective approach to increase the PEGylation efficiency which therefore enhances the stability of therapeutic enzymes. Furthermore, the PEGylated PpADI M36 represents a highly attractive candidate for the treatment of arginine-auxotrophic tumors.