Considering that data searched from the search engine is not comprehensive, and the inconsistencies between desired results and received results are inevitable, a more effective search tool called Distributed Document Clustering for Search Engine (DDCSE) is proposed in this paper. In the DDCSE, the utilizing of distributed clustering and several search engines is used to categorize the results, in order to feedback a set of better refined results. Experiments show that a significant improvement is achieved via the distribution document clustering, so as to refine the results and reduce the time used to filter out irrelevant data for the search engines.