The performance of an air gap membrane distillation (AGMD) system for water desalination was optimized and compared using Taguchi orthogonal design arrays and the response surface methodology (RSM) Face-Centered Central Composite Design. The variables considered in the optimization are feed flow rate, feed temperature, coolant temperature, coolant flow rate, and air gap width. Additionally, a quadratic RS-model between response and the operating parameters was developed. Analysis of variance was then use to analyze the model and the significant effect of each operating parameters on flux. Results showed that the feed temperature and the air gap width are the most influential factors controlling the permeate production. The sensitivity of the permeate flux to coolant flow rate is marginal compared to other factors. From Taguchi technique, a maximum flux of 76.046 kg/m² h was obtained at optimum conditions. The RSM approach produced a maximum flux of 76.998 kg/m² h at optimum conditions. Although both techniques predicted very similar results, the RSM gives better predictions of the influence of operating parameters on response, and it is a better tool for the optimization of AGMD system compared to Taguchi methodology.