A multi-objective optimization method is proposed to overcome the shortcomings of the traditional method which establish the multi-body dynamic model frequently. A multi-body dynamic model of a McPherson front suspension with composite transverse leaf spring is developed in ADAMS/Car. The parallel travel simulation result shows the toe angle, camber angle, kingpin inclination angle, and caster angle deviate from design targets. To converge to the design value and reduce the variation range, the influences of suspension hard points on the kinematic parameters are studied. Eight design variables are identified out of 36 hard point coordinates according to their sensitivity to the target. The root mean square values of the kinematic parameters are set as the optimization objectives. The quadratic polynomial is employed to build the response surface of the kinematic parameters to design variables. The Pareto optimal solution set is obtained using the NSGA-II algorithm. The optimization merges the gap between the initial values and the design targets from [0.3066, 0.3425, 12.6201, 4.7352] to [0.2936, 0.3463, 13.1535, 3.9588] with variation range of [89.42%, 29.46%, 12.05%, 15.77%]. It is shown the method can effectively improve the precision of the dynamic model and lays a foundation for the research of suspension handling and stability performance.