Abstract: Aiming at the existing problems of trivial and irregular operations in tire's reverse modeling, this paper proposes a tread pattern's hierarchical modeling method based on tire's reverse features. Firstly, the tread pattern is broken up into hierarchical knowledge elements in view of its formal features. Secondly, according to the feature-driven mechanism, feature points in different levels of knowledge elements are extracted from the point cloud, and then the feature points are processed to recognize the corresponding knowledge element. Finally, the process from knowledge elements to the model can be realized. Furthermore, in order to apply this method to the application, we developed an intelligent tread pattern reverse-modeling system, which integrates database technology and uses CATIA/CAA as the secondary development platform. The analysis of examples shows that the system can improve the efficiency of the tire's reverse modeling and standardize the modeling process.