The solvent sublation technique was applied to remove water-soluble dyes from aqueous solutions. The aim of this work is computer simulation of the solvent sublation process in prediction the amount of removed dyes. In this study, mathematical statistic methods were used to define the process variables that have the real impact on the solvent sublation process. The STAR system mathematical tools were used to find a proper model that govern the change of dye residual concentration over time, as well as to carry out parametric identification. The simulation results show that the models have a good performance in the simulation and prediction of the cationic and anionic dyes removal from aqueous solutions. The results can be used to optimize the solvent sublation process as a technique of wastewater treatment.