摘要:The purpose of this paper is to improve the banks deposits’ management process that will ultimately result in the collection of deposits at lowest costs. In order to achieve the aforementioned purpose, the authors have derived model using statistical and mathematical methods. Statistical method of double logarithmic regression was applied in order to determine the relations between the costs and amount of funds collected from each deposit product, while in the process of deposit assortment optimization nonlinear programming was applied, as the correlation between costs and the amount of collected funds from the sale of deposits products is nonlinear what is shown in the selected regression model too. In order to present the usefulness and wide possibilities of its use, the deposits management model was employed on the example of one large bank that operates in the Croatian banking market. The model is applicable both to the retail sector and corporate sector deposits. Information derived from the model application help banks management in making appropriate decisions regarding planning and managing the sales of deposits and, along with other existing instruments, complements the database of banks management information system.