In a paper, Geetha, Vanathi, and Poonthalir [1] suggested a model for the MDVRP. However, we find that authors ignore the relationship between the two types of proposed binary decision variables, and it leads to the wrong integer programming model for the MDVRP. We give an illustrative example to identify the problem and suggest new constraints to overcome. Also, authors should include the distances between depots and customers in the objective function.