Color is one of the major remaining contaminants in the palm oil mill effluent (POME) following the conventional treatment of POME. The removal of color from POME using adsorption on activated carbon was investigated. The adsorption experimental design was performed using the standard response surface method (RSM) design that is central composite design to determine the optimum process variables for color removal by using the Design-Expert software (version 7.0. Stat-Ease, trial version). Besides obtaining optimum values, RSM also has the advantage of studying the interaction between various experimental parameters compared to one-factor-at-a-time. The equilibrium experimental data were analyzed by Langmuir and Freundlich isotherms. The statistical analysis showed that the quadratic model as well as the model terms was significant. The model had very low probability value (0.0003). The R 2 for the model was 0.9184 and the adjusted R 2 was 0.8380. The validation of the model showed experimental value and predicted value of 0.124 and 0.106, respectively. The optimum conditions suggested by the model for the process variable were 87.9 min, 4.05 and 7.86 g, for time, pH and granular activated carbon dose, respectively. The maximum removal obtained at these conditions was 89.95 %. The adsorption isotherm data were fitted well to the Langmuir isotherm compared to Freundlich isotherm with R 2 value of 0.850 for the former and 0.273 for the later.