Integrating XML documents in data warehouse is a major issue for decisional data processing and business intelligence. Indeed this type of data is increasingly being used in organisations’ information system. But the current warehousing systems do not manage documents as they do for extracted data from relational databases. We have therefore developed a multidimensional model based on the Unified Modeling Language (UML), to describe an XML Document Warehouse (XDW). The warehouse diagram obtained is a Star schema ( StarCD) which fact represents the documents class to be analyzed, and the dimensions correspond to analysis criteria extracted from the structure of the documents. The standard XQuery language can express queries on XML documents, but it is not suitable for analyzing a warehouse as its syntax is too complex for a non IT specialist. This paper presents a new language aimed at decision-makers and allows applying OLAP queries on a XDW described by a StarCD.