BACKGROUND Metabolic engineering aims at channeling the metabolic fluxes towards a desired compound. An important strategy to achieve this is the modification of the expression level of specific genes. Several methods for the modification or the replacement of promoters have been proposed, but most of them involve time-consuming screening steps. We describe here a novel optimized method for the insertion of constitutive promoters (referred to as "promoter knock-in") whose strength can be compared with the native promoter by applying a promoter strength predictive (PSP) model. RESULTS Our method was successfully applied to fine tune the ppc gene of Escherichia coli. While developing the promoter knock-in methodology, we showed the importance of conserving the natural leader region containing the ribosome binding site (RBS) of the gene of interest and of eliminating upstream regulatory elements (transcription factor binding sites). The gene expression was down regulated instead of up regulated when the natural RBS was not conserved and when the upstream regulatory elements were eliminated. Next, three different promoter knock-ins were created for the ppc gene selecting three different artificial promoters. The measured constitutive expression of the ppc gene in these knock-ins reflected the relative strength of the different promoters as predicted by the PSP model. The applicability of our PSP model and promoter knock-in methodology was further demonstrated by showing that the constitutivity and the relative levels of expression were independent of the genetic background (comparing wild-type and mutant E. coli strains). No differences were observed during scaling up from shake flask to bioreactor-scale, confirming that the obtained expression was independent of environmental conditions. CONCLUSIONS We are proposing a novel methodology for obtaining appropriate levels of expression of genes of interest, based on the prediction of the relative strength of selected synthetic promoters combined with an optimized promoter knock-in strategy. The obtained expression levels are independent of the genetic background and scale conditions. The method constitutes therefore a valuable addition to the genetic toolbox for the metabolic engineering of E. coli.