Abstract: Even XML is used as a popular data exchange standard over the Internet and Intranet. Because of adding tags to every different semantic content unit. Its space expansion makes the transmitting and storing of XML data very expensive in terms of resources. After compressed, XML's size is much smaller, but how to evaluate query directly based on the compressed data still requires us to do some work. An XML index structure Interval\++ on compressed data is proposed, which is the result from revert arithmetic compression. Queries as the form of //element\-1/element\-2/…/elment\-m can be evaluated efficiently using Interval\++.