SUMMARY
In this paper, we propose a variable review period model using dynamic programming for order crossover problem. The proposed model is compared to an existing variable review period model and classical review period in terms of minimum inventory cost. The simulation is applied with six scenarios and sensitivity analysis is also done. The results show a variable review period model performs smaller inventory cost for small variation of lead-time compared to classical review period model. On the other hand, classical periodic review model is sensitive with the changes in the variation of demand distribution and service level. The proposed model gives the better solution compared to variable review period model for the six scenarios with the inventory cost saving as 11-42%. The sensitivity analysis for a variable review period model and a proposed model are applied and compared. From sensitivity analysis, it can be concluded variable review period with dynamic programming gives the best solution in terms of the changes of holding cost and ordering cost. It is also shown the smaller holding cost, the smaller percentage of inventory cost savings (between variable review model and the proposed model).