A substantial body of empirical research now supports the importance of client and therapist characteristics as well as the therapeutic relationship between them in predicting therapeutic change. Recent studies have increasingly recognised Virtual Reality (VR) and Augmented Reality (AR) as potentially effective technologies in the treatment of specific phobias. Despite these results, there is little theoretical and empirical understanding of how technologies actually support the users (i.e., client and therapist) in their respective but closely related activities; how they are actually used; and how they modify the therapeutic process. This article reviews the state of the art related to both the Human–Computer Interactions issues and the clinical aspects of the therapeutic process in both technology-mediated and non-mediated contexts, and proposes the theoretical model of the VR/AR-Mediated Therapeutic Process, which integrates the concepts from both fields. The applications and relevancy of this model are also discussed.