An iterative approach is proposed to predict the optimal concentration profiles in an adsorption process and to simultaneously obtain parameters for a multicomponent isotherm by selecting representative liquid compositions for experimental measurements. The proposed approach was demonstrated for xylene adsorption on zeolite 13X that has undergone a barium ion exchange. Breakthrough experiments were performed at liquid compositions predicted by the process model, which was refined in an iterative manner with the experimental data. Robustness of the isotherm parameters for model uncertainty and variations in the process design were also analyzed. (C) 2017 Elsevier Ltd. All rights reserved.