The importance of placebo responses for the treatment of various medical conditions has increasingly been recognized, whereas knowledge and systematic application in clinical settings are still sparse. One possible application for placebo responses in pharmacotherapy is given by learning paradigms, such as behaviorally conditioned immunosuppression, aiming at drug dose reduction while maintaining therapeutic efficacy of drug treatment. In an established learning paradigm of conditioned taste aversion/avoidance (CTA) in both, rats and humans, respectively, a novel-tasting drinking solution (conditioned stimulus, CS) is paired with an injection of the immunosuppressive drug cyclosporine A (CsA) as unconditioned stimulus (US). The conditioned response, evoked by re-presenting the CS alone at a later time, is reflected by avoidance behavior of consuming the solution (conditioned taste aversion; CTA) and a diminished interleukin (IL)-2 and interferon (IFN)-γ cytokine production as well as mRNA expression of rat splenic T cells or human peripheral T lymphocytes, closely mimicking the immunosuppressive effects of CsA. However, due to unreinforced CS-re-exposure conditioned responses progressively decreases over time (extinction), reflecting a considerable challenge for potential clinical applications of this learned immunosuppression. The present article discusses and critically reviews actual approaches, applications but also limitations of learning paradigms in immune pharmacotherapy.