Continuous fixed-bed column adsorption for wastewater treatment provides the most practical application to reduce water pollution by heavy metal ions. The objective of this work was to develop a new statistical approach capable of simulating the dynamic behavior of the fixed-bed adsorption. A statistical model (normal model), based on normal distributions, was deduced and experimentally validated in a Bayesian framework. The model was applied to describe the adsorption mechanism of Cr(VI) onto chitosan flakes, obtained from the waste of crustacean industry. The continuous flow of Cr(VI) was carried out in a glass column, at different operating conditions, determining the breakthrough curve and the breakthrough point time (tb). The proposed normal model provided a good description of the column adsorption with fitting error comparable to that of the widely used Bohart-Adams logistic model. In all cases, the tb was into or near to model credibility interval. Therefore, the model with the Bayesian approach represents a good option for the design of adsorption columns for wastewater treatment.