Abstract: In order to study the impact of censored demand information on retailers' optimal decision-making, a data-driven distributionally robust newsboy pricing and inventory joint decision model is established with the objective of minimizing the total cost including inventory holding costs and stockout penalty cost. Firstly, the censored demand information is introduced into the distribution uncertainty set, and Cox regression is used to estimate the censored value to compute the parameter of the demand uncertainty set. The original problem is transformed into a dual problem and solved by using Rsome analysis tool. Then, the efficiency of the distributionally robust optimization method proposed in this paper is studied at different parameter levels through random numerical experiments. The results show that compared with the common optimization algorithms (robust optimization and stochastic optimization), the proposed method saves an average cost of 23.74% and 16.32%, respectively, and the average cost difference between the proposed method and the theoretical optimal case is narrowed to 21.41%. Finally, the effectiveness of the distributionally robust optimization method proposed are further verified based on a real case data.