Abstract: Evaluating the performance of Decision-Making Units (DMUs) is a very important subject in Data Envelopment Analysis (DEA) and DEA does this by assigning a relative efficiency score to each DMU. In spite of the great importance of this subject, the majority of the methods presented in the different studies have not attempted to evaluate the performance of the inefficient DMUs beyond the relative efficiency scores attained from the standard DEA models. In this study, the minimum distance of each inefficient DMU from the efficient frontier is obtained by a branch-and-bound algorithm and then these distances are used as a criterion for evaluating the performance of the inefficient DMUs. It is known that using this method for evaluation is more realistic compared to other methods and also this method is able to remove the existing difficulties in some models, say, the CCR and BCC models. Also, this study proposes the use of the minimum distance of an inefficient DMU from the efficient frontier in order to obtain the shortest projection from the evaluated DMU to the efficient frontier, thus allowing an inefficient DMU to find the easiest way to improve its efficiency.