The reactive extraction of glycolic acid (GA) using tri-n-octylamine (TOA) in a mixture of inert diluent and modifier was modeled by response surface methodology (RSM), which was found to be the most preferred method for nonparametric modeling in chemical engineering. Differential evolution (DE), a bio-inspired evolutionary-based algorithm, was employed to determine global optimum parameters, due to the limitations of traditional optimization methods. The degree of extraction (%Y) was modeled and optimized considering four independent design variables: (i) GA concentration (C-in), (ii) TOA composition (C-TOA(0)), (iii) modifier composition (M), and (iv) equilibrium temperature (T). The RSM model correlates design variables to response with R-2 = 0.98. At optimum conditions (C-in = 0.1707 mol/L, C-TOA(0) = 22.31 (%v/v), M = 73.28 (%v/v), and T = 23 +/- 0.5 degrees C) offered by the RSM model and DE, the model predicted and experimental values of %Y were found to be 94.95% and 91.83%, respectively.