One of the most fundamental item retrieval problems associated with warehousing is the simple “order-picking problem”, the process of retrieving items from stocking locations in a warehouse. A customer’s order is converted into a pick list, which specifies the location of each type of items. A picker then travels through the warehouse according to a routing policy, picking the items from stock. The critical issues in the order-picking problem are routing policies and storage policies of items. In this paper, items are stored basically on the basis of the order popularity. Through the development of analytical models, we compare the performance of three routing policies, namely traversal, return, and midpoint policies, in terms of the expected travel distance. Three policies are also tested varying the parameter value of the ABC curve, the number of picks in one pick list, and the number and length of stocking aisles. Simulation results are presented confirming the accuracy of the analytical models.