A method for analysing the vehicle-bridge interaction system with enhanced objectivity is proposed in the paper,which considers the time-variant and random characteristics and allows finding the power spectral den-sities(PSDs)of the system responses directly from the PSD of track irregularity.The pseudo-excitation method is adopted in the proposed framework,where the vehicle is modelled as a rigid body and the bridge is modelled using the finite element method.The vertical and lateral wheel-rail pseudo-excitations are established assuming the wheel and rail have the same displacement and using the sim-plified Kalker creep theory,respectively.The power spectrum function of vehicle and bridge responses is cal-culated by history integral.Based on the dynamic respon-ses from the deterministic and random analyses of the interaction system,and the probability density functions for three safety factors(derailment coefficient,wheel unload-ing rate,and lateral wheel axle force)are obtained,and the probabilities of the safety factors exceeding the given limits are calculated.The proposed method is validated by Monte Carlo simulations using a case study of a high-speed train running over a bridge with five simply supported spans and four piers.