The Multi-Depot Vehicle Routing Problem (MDVRP) is an extension of the classical Vehicle Routing Problem (VRP) in which vehicles start from multiple depots to visit customers and return to their depots of origin at the end of their tours. The objective of MDVRP is to minimize the total length of the routes, same as the other types of VRP. In this paper a solution methodology based on Ant Colony Optimization (ACO) inspired by the observation of real ants is proposed to solve MDVRP. ACO includes many algorithms which are many diverse variants of the basic principle. We presented the computational implementations of our model for different benchmark MDVRP problem sets. Besides we also recognized the effects to the solutions of all the parameter changes of Ant Colony Optimization and the effects of angle changes on candidate list in detail.