The multi-objective collaborative optimization method based on NSGA-II is proposed for multi-objective multidisciplinary design optimization problems with multi-objective subsystems. The evolutionary process in which the individuals meet the feasibility is irrelevant to the target point allocated by the system level in multi-objective subsystem optimization. In view of this characteristic, a method for producing initial population with good feasibility and diversity is proposed to improve the calculation efficiency and accuracy of multi-objective subsystem optimization. In the process of NSGA-II non-dominated sorting, the effect of the interdisciplinary incompatibility function differs from the physical objectives, which increases the difficulty in sorting the individuals. To avoid this problem, a strategy of transforming the incompatibility function into a disciplinary constraint is presented in the multi-objective subsystem optimization. Finally, the engineering example shows the effectiveness of the proposed approaches.