Abstract: To address the complexity of multi-warehouse replenishment decisions and the difficulty of jointly optimizing inventory cost and service level in the distributed warehousing system of medical consumables enterprises, an inventory optimization method for multi-warehouse collaborative replenishment is proposed. Based on this, an inventory optimization model is developed with the objective of minimizing inventory holding cost, shortage cost, and transportation cost, while incorporating practical operational constraints such as product shelf-life, batch management, transportation mode economy, and replenishment quantity limits. In the algorithm design, a two-stage hybrid algorithm combining particle swarm optimization (PSO) and simulated annealing (SA) is proposed. On the basis of the PSO algorithm, global search is first performed to obtain high-quality initial solutions. Then, the probabilistic jumping mechanism of the SA algorithm is introduced to escape local optima, enabling the coordinated optimization of multi-period replenishment quantities and transportation schemes. Meanwhile, a rolling horizon mechanism is incorporated to dynamically generate replenishment plans for the next four periods, thereby improving the adaptability of replenishment decision-making. Case study results show that, compared with the enterprise′s current replenishment strategy, the proposed optimization scheme reduces the total operating cost by approximately 39.23%, including reductions of 38.60% in inventory holding cost, 62.31% in transportation cost, and 39.43% in shortage cost. The results demonstrate that the proposed method can effectively support replenishment decision-making in multi-warehouse environments and provide a feasible decision-support approach for supply chain inventory optimization in medical consumables enterprises.