According to the working characteristic of hybrid electric vehicle (HEV) in different conditions, the multidisciplinary design optimization (MDO) method in parameter optimization of power system of HEV under uncertainty was studied. To estimate the impact of uncertain input parameters on objectives of drive power system by performing the Monte Carlo simulation. The aggregate preference function of objectives and corresponding constraint equation were constructed to satisfy the designers’ preference and engineering design. The power system parameters were optimized by using analytical target cascading (ATC) method which could hierarchically decompose and collaboratively optimize the system. The results showed that the values of vehicle’s performance and the power system design could be optimized currently.