In this paper, the implications of the so-called optimized NF membrane properties on the economic assessment as well as the process performance of NF membrane system will be discussed. The predictive model for NF membranes such as the Donnan-steric-pore model has been used in conjunction with an economic assessment model such as the Verbenne cost model to study the impact of various NF membrane parameters on the capital and operating cost of NF membrane system. The parameters studied include the membrane characteristics such as pore size, charge density, effective thickness and the system characteristics such as applied pressure and concentrations. The optimal conditions for minimum operation cost and optimal operating pressure were then determined. Detailed cost analysis reveals both capital and operating cost were not affected by the variation of charge density. However the charge density will significantly affect the rejection of the membranes. On the other hand, decreasing the membrane pore size will increase costs. It was found that up to 30% of the capital and operating cost in recovering sodium sulphate solution can be saved by using looser membrane compared to the tight membrane structure. Some efforts to experimentally produce NF membranes with optimum properties will be briefly illustrated.