Surface response methodology was involved in the optimization of Cr(VI) adsorption upon chitosan flakes against the process parameters pH, adsorbent dose and initial Cr(VI) concentration. The effects of these factors were studied in the ranges 1.5-9.5, 1.8-24.2 gl(-1) and 15-95 mg l(-1), respectively. A predictive quadratic model was constructed by variance analysis of data obtained from a total of 20 experimental runs with three replicates each. Maximum removal was attained from a solution as concentrated as 30 ppm at pH 3 with an adsorbent dosage of 13 g l(-1). The adsorption capacity of chitosan flakes was determined as 22.09 mg g(-1) at these specified conditions. However, the adsorption capacity was recorded as high as 102 mg g(-1) for 100 mg l(-1) initial concentration. Out of Langmuir, Freundlich and Dubinin-Radushkevich isotherm models, adsorption data was best described by Langmuir isotherm with 0.99 consistency. The process kinetics was evaluated by pseudo-first, pseudo-second order and infra-particle diffusion models. Pseudo-second order kinetic model exhibited the highest correlation with data. The results showed that both monolayer adsorption and infra-particle diffusion mechanisms limited the rate of Cr(VI) adsorption. Thermodynamic parameters revealed the feasibility, spontaneity and exothermic nature of adsorption. (C) 2009 Elsevier B.V. All rights reserved.