It is difficult to do the fault diagnosis on the modern car engines which have high technology and complex structures. In this study, a case-based-reasoning (CBR) based automobile engine intelligent fault diagnosis system was proposed against this problem. The system’s structure and its mechanism of fault diagnosis were introduced. The key techniques to implement the system were analyzed, including the case establishment, the case search, the case learning and the maintenance of case library. The proposed system gave a new way to establish an efficient automobile engine fault diagnosis system.