Long-span railway sea-crossing bridges are subjected to strong winds and high waves. Due to variations in structural features, including stiffness and size across diverse orientations, the dynamic response of vehicles and bridges can exhibit significant differences under distinct wind and wave directions. Consequently, a novel approach for analyzing bridge and vehicle dynamic performance called the directional wind-wave-vehicle-bridge (DWWVB) model has been developed. To acquire the statistical characterization of the correlation between wind and wave conditions, a three-dimensional joint probability distribution of wind and wave characteristics derived from collected data at the bridge site is established. The wind loads are derived from wind tunnel tests and computational fluid dynamics simulations. The nonlinear wave loads are solved utilizing the hydrodynamic solver ANSYS-AQWA. Based on the finite element bridge model and a 23-degree-of-freedom mass-spring-damper vehicle model, the aerodynamic nonlinearity and initial geometric nonlinearity are considered during the iteration process. The Newmark-β method is employed to solve the dynamic response of bridges and vehicles. The calculation results reveal that the most unfavorable response occurs when the wind direction is 45° and the wave direction is 90°. The directional analysis provides valuable insights for selecting optimal bridge locations and reasonable design environmental parameters.