The application of response surface methodology (RSM) in preparation and optimization of membranes is important in order to reduce the effort and time needed to achieving an optimum performance. RSM was used to develop an optimum polyethersulfone (PES)/ZnO mixed matrix (MM) membrane for humic acid removal. The MMs were synthesized by dispersing various amounts of hydrophilic ZnO nanoparticles (NPs) into a solution containing PES, polyvinylpyrrolidone (PVP) and dimethylacetamide (DMAc). Flat sheet MM membranes were prepared via the phase inversion method using the central composite design (CCD). The effects of four preparation parameters, such as PES, ZnO, PVP weight percentages and solvent evaporation time, were investigated. Pure water flux (PWF), humic acid flux (HAF) and humic acid rejection (HAR) were selected as a model responses. It was shown that PES and PVP were mainly affected on both PWF and HAF. Furthermore, the interaction effect between PES and ZnO-NPs shows a significant effect on PWF, while the quadratic effects of both solvent’s evaporation time and ZnO-NPs weight percentage coupled with the interaction effect between PES and PVP weight percentage shows the most significant parameters that affects HAR. The optimization method was subjected to maximize all of the PWF, HAF and HAR. It was also determined that the optimized membrane can be synthesized from a solution containing 17.25 wt% PES, 3.62 wt% ZnO and 3.75 wt% PVP with 15 s of solvent evaporation time. The optimum values of PWF, HAF and HAR were 222.3 (L/m2 h), 94.7 (L/m2 h), and 96.34%, respectively. Thus, it can be concluded that the CCD technique is capable of optimizing PES-ZnO membrane performance.