In this paper, experimental modal analysis (EMA) is carried out using the actual structure of an autonomous bus chassis, and the corresponding finite element model (FEM) is verified by comparing the modal parameters obtained from EMA and finite element analysis (FEA). In EMA, impulse force is applied to the structure through an impact hammer, which has the characteristics of broadband containing frequency content around the excited structure modes of interest. We use the contact condition to replace the bolts and rivets of the chassis structure through FEA, and then to reduce the number of finite elements and improve the computational efficiency. The reliability of FEM is judged by comparing modal-estimation results obtained from EMA and FEA. The error of natural frequencies and the modal assurance criterion between modal shapes through EMA and FEA are used to verify the consistency of modal parameters. On the basis of comparison results to determine the reliability of FEM, the bending stiffness, torsional stiffness, and strength of the chassis structure are evaluated by applying corresponding boundary and loading conditions, and these values are served as safety indicators for the chassis structure.