The present study aimed to route vehicles from depot in two-level supply chain with a meta-heuristic algorithm. This study is an analytical type of research. The issue of routing which includes time, travel duration dependent on the day, has first been introduced by Mandraki and Duskin. They proposed the issue of time-dependent Travelling Salesman Problem (TSP), such that their simplified issue was VRP issue and service should be provided to all customers, and each customer should only be visited by one vehicle. Mathematical modeling has been used as research tool, and in the examined problem, there are a number of demand points with simultaneous delivery and return. This demand should pass the depot which should be selected from among candid points, and it should be determined that construction of depot in each of these locations has a specific fixed cost. In order to send service from depot to demand points, vehicles are used each of which has its special carrying capacity, fixed using costs, and overhead carrying costs. Given that solving the problem is time-consuming, MOPSO meta-heuristic algorithm has been used in order to solve the problem. The obtained results have been fully presented, and in different repetitions it was observed that the second vehicle has the highest load carrying.