Scholars have proposed many methods of Data Envelopment Analysis (DEA) for re-ranking already. However, in many re-ranking methods, some cannot guarantee to effectively re-rank the decision making units (DMUs) which have the same efficiency. In this research, we would propose a method that can guarantee to be completely re-ranking. Through further investigating and originally research, the Expanded Feasible Region Method for re-ranking is proposed by this research. This method not only has the function of re-ranking, it can also attest the thought that decreasing number of input items or output items can increase the discrimination of efficiency. To proof the efficiency of the method mentioned in this research, we will not only give examples to explain the method of re-ranking, but also proposes one theorem and two corollaries to prove it, showing the accuracy of the concept adduced in this research.