This study aimed to explore influence on the design of vehicle front shape for a SUV by using pedestrian and bicyclist. The injuries of head, upper leg and lower leg based on 50th human model are considered as design response.A multiobjective optimization approach has been developed to design and optimize the vehicle front shape for minimizing injury risks of pedestrian and bicyclist in this study. The numerical model of pedestrian and bicyclist impact vehicle was constructed first with MADYMO program and stiffness curves of vehicle were validated with physical test. Then, the optimal Latin hypercube sampling (OLHS) method was adopted for design of experiment (DOE) and the surrogate model was constructed through Radial basis function (RBF). The optimal problems involving in a number of objectives were solved by the multiobjective particle swarm optimization (MOPSO) algorithm in this study.