•A deep reinforcement learning algorithm is proposed for multi-item retrieval in the PBS system.•A compact integer programming model is built to evaluate the solution quality.•A conversion algorithm is proposed to handle simultaneous movement.•A decomposition framework is designed for large-scale instances.•The effect of several factors is investigated to deduce managerial insights. Nowadays, fast delivery services have created the need for high-density warehouses. The puzzle-based storage system is a practical way to enhance the storage density, however, facing difficulties in the retrieval process. In this work, a deep reinforcement learning algorithm, specifically the Double&Dueling Deep Q Network, is developed to solve the multi-item retrieval problem in the system with general settings, where multiple desired items, escorts, and I/O points are placed randomly. Additionally, we propose a general compact integer programming model to evaluate the solution quality. Extensive numerical experiments demonstrate that the reinforcement learning approach can yield high-quality solutions and outperforms three related state-of-the-art heuristic algorithms. Furthermore, a conversion algorithm and a decomposition framework are proposed to handle simultaneous movement and large-scale instances respectively, thus improving the applicability of the PBS system.