To optimally allocate arrival and departure passenger trains among passenger stations in a railway terminal that connects high-speed and intercity passenger railway lines, a multiple-objective nonlinear mixed integer programming model was proposed. In the model, decision variables include distribution of arrival and departure passenger flows, and routes of arrival, departure and through passenger flows; constraints are the carrying capacity of stations and railway lines, and demand of passenger transportation; and objectives are the costs of all passengers in the terminal. An exact algorithm was designed by combining enumeration method, multiple shortest algorithm and linear programming in consideration of that a railway terminal connects a few routes. An example demonstrates that the proposed method could quantitatively solve the problem of optimally allocating arrival and departure passenger trains among 3 passenger stations in a railway terminal.