The success of simulated moving bed (SMB) chromatography for separating difficult or nearly impossible mixture has drawn strong interest in improving its performance in recent years. In this work, experimentally verified mathematical model was used to find the best possible operating conditions for continuous separation of 1,1'-bi-2-naphtol racemate under specified product quality requirement. Equilibrium dispersive model for SMB systems coupled with non-linear equilibrium isotherm and lumped kinetic approximation constitute the mathematical model in simulating the system behavior. Single objective as well as multi-objective optimization with diverse objectives was carried out using a state-of-the-art AI-based non-traditional optimization technique, elitist non-dominated sorting genetic algorithm with jumping genes. The optimization results showed that significant improvement could be achieved depending on desired objectives. Equilibrium triangle theory was used to explain the optimization results. (C) 2005 Elsevier B.V. All rights reserved.