<p>This paper presents a study of the impact that collaboration in inventory has on distribution costs in a network of one supplier and multiple customers. Freight transport is widely studied due to its impact on logistics costs and service levels of enterprises. Looking for network configurations that improve some of these aspects will always be attractive to both academics and businesses. The collaborative distribution system was proposed using the IRP model, and it was compared with the scenarios in which the both supplier or customers optimize their inventory levels. For the IRP model solution a genetic algorithm was used, from which it was obtained a collaborative distribution system with a better performance than if the optimization is made independently.</p>