Benchmarking requires an effective methodology for finding the best performer, which entails an evaluation of the relative efficiencies of competitors in terms of multiple input and output factors. To identify the best performer, Data Envelopment Analysis (DEA) has been popularly used. However, the conventional DEA has some deficiencies with respect to its use for benchmarking. First, the reference set of an inefficient DMU often has multiple efficient DMUs. Second, it might be quite impossible for an inefficient DMU to achieve its target’s efficiency in a single step, especially when the target is far removed from the DMU. To overcome these deficiencies of conventional DEA, we propose a new stepwise benchmarking method using DEA, which enables inefficient DMUs to select the more appropriate benchmarking DMU based on the similarity.