In recent years, the stochastic dynamic analysis of vehicle-bridge (VB) coupled systems with the random system parameters has attracted much attention. Nevertheless, a little work has yet been carried out to establish the stochastic dynamic model of vehicle-track-bridge (VTB) systems and the systematical evaluation on stochastic dynamic properties of VTB systems is still very limited. In this regard, an efficient approach for the stochastic dynamic analysis of time-dependent VTB systems based on the probability density evolution method (PDEM) is proposed in this study with the stochastic influences of the track structure considered. Firstly, the VTB stochastic model with the stochastic track structure is established and its stochastic motion equations are derived based on the principle of virtual work. Then, the representative parameter sample point sets of the random system parameters can be chosen and their initial probability distribution can be determined by using the number theory method (NTM). An efficient procedure for the stochastic dynamic analysis of the VTB system is presented based on PDEM. A real VTB system constructed in China is taken as an example to examine the feasibility and reliability of the proposed stochastic model and evaluation approach. The parametric study is conducted to investigate the effects of random system parameters on the dynamic responses of the VTB system. The accuracy and efficiency of the proposed approach are verified and some observations of the effects of the stochastic track structure are made.