Designing new refrigerants is a highly advanced field, where it is essential to create efficient refrigerants to meet the growing demand. However, the current approach of relying on experts’ experience or existing refrigerant databases may not result in the best refrigerant. In this paper, a refrigerant selection method is proposed that combines the selection of refrigerants with the optimization of the refrigeration process, thus allowing for the selection of the best refrigerant based on the requirements of the refrigeration process for single and cascade system. The paper proposes the combination of properties databases of different refrigerants in the refrigeration process model, enabling both operating conditions and refrigerants to be decision variables. The coefficient of refrigeration performance (COP) serves as the objective function, with 285 compounds as candidate refrigerants. By optimizing this model, both the optimal refrigerant and operating conditions can be obtained simultaneously. However, due to the high complexity of the model, the authors propose an algorithm based on decomposition to efficiently solve the mixed integer nonlinear program, as it provides the best performance. Results demonstrate that the proposed algorithm can achieve the optimization efficiently. Besides, the more complex the problem, the more time it can save.