We address three issues that are important when considering somatic gene therapy approaches to osteoarthritis (OA) and related syndromes. First, only those diseases for which a precise molecular etiology has been established should be contemplated for somatic gene therapy. Second, DNA sequences should be identified that restrict expression of correcting genes to chondrocytes; we discuss the use of transgenic mice to identify such sequences. Third, we emphasize the usefulness of establishing animal models that mimic human OA syndromes by genetic manipulations. These transgenic models should be essential for testing gene therapy approaches in vivo.