Abstract: In this study, researchers consider random disruption risks in designing a reliable distribution network model. Researchers consider two types’ distribution centers, one that is unreliable and another that is reliable. Researchers develop the three echelon of logistic model considering manufacturing plant, distribution centers and customers. Researchers consider the distribution centers are subject to probability of disruption. If distribution centers fail, customer may assign to other reliable distribution centers and incur excessive transportation cost, handling cost and opening cost of facility. Researchers consider the case where facilities are subject to probability failure. If facilities fail, customer may assign to other facilities and incur excessive transportation cost and opening cost of the reliable facility. Researchers formulate this as a mixed integer programming model and determine optimal facility location, as well as customer assignment strategies, allowing the facilities to fail under disruption probabilities consideration. The goal is to minimize the sum of initial opening cost of distribution center, transportation cost from manufacturing plant, handling cost at distribution centers and expected customer transportation cost under primary and backup assignment.