An important warehouse operation decision concerns the definition of a storage location policy, including the definition of functional areas and the identification of each stock item location in the warehouse. Empirical approaches are often used. However, they do not guarantee that the lowest cost materials-handling solution is achieved. This paper presents a clustering approach to address the assignment of stock items to particular warehouse functional areas. A linear programming model is adapted to a case study and its performance is assessed through the use of real data. Model is solved using LINGO 9.0 mixed integer commercial solver. The output of this model may also be taken as a starting point for further detailed warehouse operation decisions, especially as far as the optimization of the order picking process is concerned.