With the advancement of electricity market reform and the promotion of demand response, more and more users are actively participating in the electricity market. However, the goal of the traditional islanding partition was the maximum power supply and the maximum network loss, ignoring the active participation of the users. Therefore, considering the importance, controllability and uncontrollability and demand response, a multi-objective model of islanding partition was proposed. The model was based on the maximum amount of power supply restoration, the maximum users’ satisfaction in the process of utilizing electricity and the maximum cost of demand response. NSGA-II algorithm was used to the model in order to obtain the multi-objective pareto frontier solutions. Finally, three groups of solutions for comparative analysis, results show that the method can expand the power supply range, increase the numbers of users in the island, maximize the utilization rate of distributed generation. Meanwhile, it can improve the users’ overall satisfaction and optimize the cost of demand response.