In this study, we developed a Joint Replenishment Problem (JRP) model; In addition, JRP was introduced through capacitated vehicle to send the parts. The model was developed as an Integer-non linear model. Cost function consists of the major ordering cost, the minor ordering cost and the holding cost. Previous studies had not considered transaction cost; therefore, we considered transaction cost, which depends on the number of different parts in the order. Demand for parts was considered as constant and the model was solved by using a genetic algorithm. For calculating variable, we developed the present algorithm in the literature. We solved the model for two conditions, with dispatch volume restriction and without dispatch volume restriction. The results showed that the dispatch volume increases both the cycle time and the total cost. The model was solved for one of the major manufacturers of SAPCO. We considered the problem with 12 strategic parts of this company. The model was investigated through the real data gathered from the SAPCO Company. Finally, for model validation, the results were compared with the traditional ordering system of SAPCO, in which the parts are ordered separately. It was shown that the proposed system has better performance than the traditional ordering system. Validation of Model was performed by the numerical example that has been solved in the literature. Keywords: Joint replenishment, Automotive parts, Genetic algorithm, Inventory.