Numerous parameters of conceptual hydrological model have an inter-constraint relationship between each other and it is difficult to choose the optimal parameters of multi-objective parameter optimization due to the influences of subjective factors of policy makers. To solve this problem, we adopted a multi-objective optimization algorithm to calibrate hydrologic model parameters and obtained a series of Pareto optimal sets of model parameters. Based on these sets, we introduced the minimum maximum regret decision theory and combined the basic theory of Pareto dominance. Then we put forward a multi-objective optimal selection principles of Pareto solutions. Taking the Zhexi watershed as the research object, we used MOSCDE to calibrate the parameters of hydrological model and compared the results with the single objective optimization results. The results indicated that the proposed method can effectively select the optimal solution and is not limited by the large scale of Pareto optimal sets. In addition, this method can also greatly reduce the amount of time.