This study focuses on the integration of intelligent documents and intelligent systems. First, the paper defines the intelligent document as an electronic document that has extra self-description information, semantically. We believe that the intelligence of the document would be an important factor that impacts the performance of information retrieval systems. Next, by exploring the development of XML, we find that the XML-based technologies already became the principle of intelligent documents. Moreover, a series of system implementations have been done in this paper, they are a data exchange system, a news publication system, an XML-based CMARC cataloging system and a WAP-based OPAC system. These experiments demonstrate the application potentials of XML in many fields, such as electronic data exchange, electronic publication, library automation and mobile information service of library.