Currently, a large number of documents are created as digital material and distributed world-wide. Digital materials are easy to publish and copy at a remarkably low cost. As a result, many documents are copied illegally, and this practice is spreading, making plagiarism a significant social issue. Therefore, the need to develop systems that detect plagiarism is very high. We have developed a new plagiarism detection method that compares documents by using approximate string matching to detect plagiarism. We have also developed a technique that reduces the computational time of the comparison method. In this paper, we demonstrate our proposed method’s usefulness through experiments and through the measuring indexes of precision and recall.