Reverse electrodialysis (RED) can be utilized for the production of renewable energy from salinity gradients. However, there are many key parameters that could influence the performance of RED. This study investigates the use RSM for development of a predictive power density (PD) and open-circuit voltage (OCV) model for the RED system. A three-factor central composite design (CCD) was used to quantify the effects of flow velocity (X1), salinity ratio (X2), and number of cell pairs (X3) towards PD and OCV. A total of 17 experimental data were fitted and ANOVA was used to validate the accuracy of the models. 3D and surface plots were created to foresee the optimal levels of each variable. It was found that flow velocity and salinity ratio have the most dominant influences on the RED performances as compared to number of cell pairs. The predicted PD and OCV values were found to be reasonably fit with the experimental data, validating the predictability of applied models. Therefore, this study suggests that CCD can be considered an effective tool for evaluating and optimizing the RED system using a minimum number of experiments.