Web usage mining is the process of applying data mining technique to the discovery of user behavior patterns based on Web log data, for various applications. In the advance of e-commerce, the importance of Web usage mining grows larger than before. As an important research field of Web usage mining, mining users navigation patterns is the fundamental approach for optimizing the frame design of Web site. In this paper, we propose an ant colony approach for users navigation patterns. Firstly, a Web site model is built. Secondly, a navigation model is established, based on ant colony approach and Web logs. Finally, we design an algorithm regarding all users as entity for mining their preferred navigation paths.