Abstract: CBR(case-based reasoning) is a problem solving technique of using previous experience and cases to find a solution to new problems, which is suitable for those areas of experience-rich. CBR methodology was applied to the fault diagnosis and maintenance decision support of aircraft due to the abundant repair experience accumulated in after-service of aircraft. According to the requirement of aircraft maintenance, the research into the method of aircraft fault diagnosis using CBR was conducted and key techniques, including the case base development, case representation, case retrieval, adaptation and retain, were investigated. The case representation was discussed in detail, and a case retrieval model of combining the character field match with the KNN(K-nearest neighbor) method was established and applied. A framework of aircraft fault diagnosis system was proposed. The prototype system was developed and an application example was given.