Abstract-This research is intended to address the cases retrieval phase related to a CAPP system for axisymmetric solid parts supported by the Case Based Reasoning (CBR). The traits of similarity between two parts were identified from the characteristics of the stock and the finished part. As a result of this work, a function of global similarity metric was obtained, and it incorporates in addition to the geometry of the part and the stock attributes: the type of material, the type of stock and the relationship of the diameter-length dimensions of the part. The results constitute an improvement to the retrieval phase of cases with regard to the retrieval of the most similar case in the repository for the execution of new technical task.