As XPath is the core of most XML query languages, the efficiency of processing XPath expression has been a main part of the cost in XML queries. However, most existing XPath processing algorithms, which don't take index structure into account, spend lots of costs on spaces and time. This paper proposes an efficient XPath query processing mechanism based on structural index that makes full use of XML structural index to quickly retrieve XML data. Coming together with the mechanism, the Compressed XPath Query Tree based on structural index is proposed, which significantly reduces many join operations. Then query algorithms are used to deal with all the structural relationship using our mechanism. Finally, experiments will run the algorithm on real XML datasets and query workloads to report the performance of our mechanism and show the efficiency compared with other mechanisms.