Optimization is an important part of the design on Unmanned Surface Vehicle (USV). In this paper, considering the rapidity, maneuverability, seakeeping and rollover resistance performance of the USV, the design variables of the optimization system of the USV have been determined a mathematical model for comprehensive optimization of USV has been established. Integrated optimization design of multi-target and multi-constrain is achieved by computer programs. However, the influence degree of each design variable are different on the final optimization results, in order to determine the degree of influence of each design variables, find out the key variables for a further optimization analysis and sensitivity studies of the design variables to optimization will be crucial. For solving this problem, a C++ program has been written by genetic algorithm and five discrete variables have been selected which are used to study the sensitivity of optimization. The results showed that different design variables have different effects on the optimization. The length of the ship and the speed of propeller have the greatest effect on the total objective function. The speed of propeller has a greater impact on both rapidity and seakeeping. The length of ship L, the molded breadth of ship B, the draft of ship T and design speed Vs have a greater sensitivity to maneuverability. Also, molded breadth B has the greatest effect on the rollover resistance.