A comprehensive optimization study was carried out to evaluate the performance of a simulated moving bed reactor (SMBR) system for an industrially important biochemical reaction-separation problem, the inversion of Sucrose and the in situ separation of the products, glucose and fructose. Two modifications of SMBR are studied, one in which non-synchronous switching is used to vary the number of columns in different sections within a global switching period (the so-called Varicol® process), while in the other the concept of distributed feed is studied in which the feed flow was distributed over the global switching period. Multi-objective optimization study is performed as it results in more meaningful solutions although the optimization for such complex processes is complicated owing to the complex interplay of relatively large number of decision variables, including continuous variables like switching time, flow rates and column length and discrete variables like number and distribution of columns in different sections. A state-of-the-art non-traditional but more versatile optimization technique based on genetic algorithm, non-dominated sorting genetic algorithm with jumping genes (NSGA-II-JG) was used. The optimization results in significant improvement for SMBR as well as its modification, Varicol and distributed feed systems, in terms of increasing productivity using less desorbent for both existing as well as at design stage. © 2004 Elsevier B.V. All rights reserved.