With the aim of fully utilizing the transport capacity of high-speed trains during non-peak periods, high-speed railway express delivery has become a new trend in the development of railway express delivery. However, the uncertain capacity of passenger trains for express delivery due to passenger flow demand poses a significant challenge to daily operations. To enhance the capability of high-speed railway express delivery in coping with uncertainty of train transport capacity, a stochastic optimization method for high-speed railway express delivery planning considering random capacity is proposed. Firstly, the uncertainty of capacity is modeled using discrete scenarios. With the precise probability distribution of the scenarios being known, the two-stage stochastic programming model(SP) is constructed to maximize the expected total revenue of high-speed railway express delivery, considering delivery time requirements and operational constraints of express delivery comprehensively. On this basis, a two-stage distributionally robust optimization model(DRO) is developed for scenarios where probability distribution information of uncertain capacity is partially known. By utilizing dual theory and box ambiguity set, the DRO model is transformed into an equivalent integer linear programming model, which is solved using the GUROBI solver. Finally, the effectiveness of the model is verified by numerical experiments based on the Nanjing-Hangzhou high-speed railway. Results show that, compared with the stochastic programming model, the DRO exhibits strong robustness and can effectively resist the impact of capacity fluctuations on the transport plan at a small cost. Moreover, it can improve the quality of solutions in the worst case and enhance the stability of high-speed railway express delivery operations in practice.