Perishable inventory systems are characterized by finite product lifetimes, uncertain demand, and the economic consequences of both shortage and spoilage. These characteristics make conventional inventory policies, such as the classical economic order quantity and basic   policies, insufficient for many practical applications. This paper develops a mathematical framework for optimal ordering policies in a periodic-review inventory system with fixed shelf life. The model incorporates stochastic demand, deterministic product expiration, ordering costs, holding costs, shortage penalties, and disposal costs. A dynamic programming formulation is presented, together with a decomposition of inventory by remaining lifetime. The resulting policy is shown to depend on the age profile of inventory rather than solely on aggregate stock. Structural properties of the optimal policy are discussed, including state dependence, order-up-to behavior, and the value of first-expiring-first-out issuing. A computational solution approach based on backward induction and value-function approximation is proposed for practical implementation. The framework provides a foundation for designing replenishment policies for food, pharmaceuticals, blood products, and other short-life products.