For all pressure driven membranes, one of the main problems which hinders the membrane practical application is the permeate flux reduction due to the solute accumulation on the membrane surface. The most popular explanation for the flux decline supported by Bhattachajee &Bhattacharya (1993), contains two mechanisms: concentration polarization(CP) and fouling. The influence of CP is noteworthy in ceramic nanofiltration system. On the one hand, CP can influence the performance of membrane separation by decreasing the retention of the molecules. On the other hand, CP could have a desirable effect which can be used for membrane surface modification. In the past three or four decades, several different models have been used to verify the existence of CP or cake-enhanced CP(CECP) effect and try to quantify it. However, all these methods or models have their own limitations. Therefore, it is essential to build a new model or adjust the constants in the empirical model according to the practical situation. The flux decline behaviour of a ceramic nanofiltration membrane in the presence of polyethylene glycols (PEGs) and silica was investigated to examine the control factor in flux decline and calculate the CP factor in the filtration. The control factor in flux decline for PEGs is CP, while for silica, both CP and fouling are important. Based on the reversibility of CP and fouling, the Gel-polarization model together with the corresponded filtration method generated the modified Gel-polarization model which is suitable for calculating the fouling resistance and the osmotic pressure on the membrane. Sherwood formula is appropriate for calculating CP factor with calibrated constants. The flux decline behaviour, as well as the CECP model developed in this work, was used to investigate the possible CP&CECP during ceramic nanofiltration for phosphate retention. CECP model based on Sherwood relation can be used to investigate the influence of the fouling layer on CP with measured permeate flux, fouling mass, and an assumed/measured porosity of the fouling layer. Based on the CECP model analysis, lower crossflow velocity and cake layer porosity, larger permeate flux and fouling mass can produce a higher CECP factor. The change of permeability in phosphate retention can be used to calculate CP factors, however, the adsorption and electroviscous effect had influence on the accuracy of the results. CECP factor is not able to be measured by the change of permeability since the unstable fouling layer can influence the discovery of permeability decrease. The presence of calcium has a serious negative impact on phosphate retention probably due to the lower electrostatic repulsion of phosphate.