ABSTRACT Spare part stock management attempts to ensure that the failed equipment items can be replaced immediately to maintain a sufficient productivity level. In maintenance, the inventory policy determination for spare parts is an important issue. The age-based preventive replacement policy may seek the least total cost for spare part replacement. Considering the criticality of equipment where it is installed, demand for a certain spare part can be categorized into critical and non-critical. The stock level for critical demand can be set to reserve certain spare parts in stock for critical equipment users. This stock policy is named the inventory rationing policy. The stock policy has a significant impact on production system performance, particularly when a machine breakdown causes a large amount of production lost. This study optimizes the joint age-based preventive replacement and inventory rationing policy in a multi-echelon spare part logistics network. In this study, there are equipment users, spare part distribution center and spare part suppliers in the multi-echelon system. Inventory policy optimization for the spare part logistics network is extremely complicated. This paper adopts the scatter search based simulation-optimization method to obtain the optimal configurations with respect to maximizing the total profit of the spare parts logistics network. An implementation example of a spare part logistics system is used to demonstrate the advantages of employing the joint policy.