As the Extendible Markup Language (XML) becoming a de facto standard for representing and exchanging data over the Internet, it is critical to be able to retrieve XML data efficiently. One way to achieve this is to use indexing, just like we index data stored in relational databases. In this paper, we present a multi-resolution structural index (MRI) method to facilitate fast retrieval of XML data. The indexing method is based on a new coding scheme that assigns unique numbers to the elements on all possible paths of the tree representing the XML document. The coding scheme is based on the DTD (Data Type Definition) of the XML file. Elements are stored in internal data structures in such a way that they can be directly accessed via the unique coding. The ancestordescendant relationships among the tree elements are easily identified to help fast process of user?s queries. Our experiments show that the MRI indexing approach is effective, both in time and space.