Abstract: To minimize the total cost of vehicle transport and to satisfy customers, it proposed a new mathematical model for multi-objective optimization of Multi-Depot collaborative vehicle routing with time windows in logistics. For the sake of this multi-objective optimization, a multi-objective evolutionary algorithm, based on decomposition, was adopted. In this algorithm, a new encoding method, which was beneficial to producing feasible individual, was presented. The efficiency of the algorithm was improved due to the perfect encoding. Finally, a test was carried out. The results show that the proposed model can solve effectively the problem of collaborative vehicle routing in logistics.