Standard enzyme engineering strategies relies on one or several amino acids permutations among the 20 amino acids (AA). The natural AA repertoire displays only limited chemical functions, which further restrain potentialities of engineering tailored proteins. To circumvent such limitation, systems have been developed over the last decades to incorporate into proteins non-canonical amino acids (ncAAs) with non-naturally encountered chemical functions [1]. The applications of ncAAs use are multiple, including protein labeling [2], protein immobilization, and in depth redesign of enzyme active sites, thus opening avenues for new catalytic opportunities [3,4]. Despite the tremendous potential of ncAAs, their use is still limited because of technical constraints. The main bottleneck consists in the poor incorporation efficiency, which can be dependent or at least related to the ncAA itself, the incorporation position and the target protein. Optimizing the incorporation system is required to overcome these limitations to efficiently produce proteins with ncAA incorporation in a more versatile way and at high production yields. Genetic code expansion is based on the reassignment of a nonsense codon to an ncAA by introducing an orthogonal amino-acyl tRNA synthetase (aaRS)/tRNA pair. In E. coli, the pEVOL system is the historical and most widely used [5]. The pUltra system allows improved incorporation efficiencies in some conditions and can be combined with the pEVOL system for the incorporation of two different ncAAs [6]. While these systems have proven their value, the incorporation efficiency remains highly variable. To go further in improvement and provide to the community a more efficient tool for ncAA incorporation, the pINS system has been developed. We focused on the expression levels of both the tRNA and the aaRS. We have demonstrated that the incorporation efficiency was increased for the three different tested ncAAs, either at the surface or surrounding the catalytic site. In addition, the incorporation position bias observed with standard systems was suppressed. The pINS system allows satisfying incorporation efficiencies with reduced ncAAs concentrations. Finally, the overall production level was increased up to 4-fold compared to pEVOL. The pINS system, making ncAA incorporation more efficient, more reliable and cheaper, should facilitate the use of ncAA in many areas of enzyme engineering.