This paper presents a model for assessing the efficiency of logistics activities in distribution centers. The DEA approach is adopted to compare the relative efficiency of distribution centers, where considered as input and output factors are warehouse floor area, field storage area, average inventory level, overhead costs, number of workers, number of orders, and total value of goods handled. The artificial neural network approach is also adopted to overcome the limitation of DEA. The 12 distribution centers of Korea Telecom are studied for the validation of the model, which results in 84.9% of learning accuracy. This model can be used to identify the inefficient factors in a distribution center and to reveal changes in the degree of efficiency over time.