Multiple sclerosis (MS) is considered a T cell-mediated autoimmune disease, and myelin proteins are the most likely candidate autoantigens. Based on experiments performed in experimental allergic encephalomyelitis (EAE), innovative immunotherapies have been developed that target either the specific trimolecular complex of encephalitogenic T cells, consisting of T-cell receptor (TCR), major histocompatibility complex (MHC; HLA in humans) class II molecule, and autoantigenic peptide, or the effector functions of these cells. To provide the basis for the transfer of these specific immunotherapies to MS, we extensively characterized the human T-cell response to one major myelin epitope, the myelin basic protein peptide (83-99). We analyzed restriction element, TCR usage and affinity, fine specificity, cytokine production, cytolytic activity, and expression of surface molecules on 41 T-cell clones (TCCs) derived from MS patients and normal controls. We demonstrate a high degree of complexity of recognition patterns as well as of functional phenotypes among T cells responding to the same epitope. In contrast to results from animal models, these findings indicate that the design of epitope-based specific immunotherapies for MS is more difficult than previously thought.