This study focuses on the adaptive mesh method for simulation of crystallization processes. Because the models of the processes are very complex, giving partial and ordinary differential, algebraic, and integral equations, it is important to get a numerical solution. The simulation is based on the method of characteristics to avoid the difficulties of numerical diffusion and stability that occur in other methods (e.g., finite difference methods for first-order derivatives) used to solve these equations. The simulation includes the kinetics of breakage as well as agglomeration. This study combines an adaptive mesh method with the method of characteristics in order to improve the accuracy and efficiency of the simulation. The proposed method can handle stiff nucleation problems of the characteristic method and conserve the crystal number and mass during mesh adaptation. The proposed method is validated through the simulation of a potassium sulfate crystallization.