The peptide p89-101 (Val-His-Phe-Phe-Lys-Asn-Ile-Val-Thr-Pro-Arg-Thr-Pro) of myelin basic protein is encephalitogenic in mice expressing H-2q and H-2s antigens. Six of 13 encephalitogen-specific T-cell clones were shown to express the variable beta-chain (V beta) 17a gene product (KJ23a+), whereas seven clones were KJ23a-. Both KJ23a+ and KJ23a- subpopulations were encephalitogenic in SJL/J mice when adoptively transferred. Depletion of KJ23a+ cells in vivo with the administration of the antibody KJ23a suppresses experimental allergic encephalomyelitis induced with KJ23a+ T-cell lines. However, experimental allergic encephalomyelitis induced with either (i) encephalitogenic peptide p89-101, (ii) intact myelin basic protein, or (iii) KJ23a- T cells reactive to p89-101 cannot be prevented with monoclonal antibody KJ23a. These data indicate that in spite of the V beta 17a gene expression in a relatively large proportion of p89-101-specific T cells, such V beta gene use is not essential for the induction of experimental allergic encephalomyelitis in SJL/J mice. These results contrast with the predominance of V beta gene use (V beta 8.2) in T cells reactive to the encephalitogenic fragment (pR1-11) in PL/J mice. One reason for this lack of dominant use of a particular T-cell receptor V beta gene family in the autoimmune response to myelin basic protein in SJL/J mice stems from the observation that two encephalitogenic epitopes exist in p89-101. KJ23a- T cells are stimulated by the deleted peptide p89-100, whereas KJ23a+ T cells are not. Thus, in the response to an encephalitogenic fragment of myelin basic protein containing two nested epitopes, at least two distinct T-cell receptor V beta genes are expressed. These distinct T-cell subpopulations can each trigger experimental allergic encephalomyelitis. These findings have implications for therapy of autoimmune disease with antibodies to the T-cell receptor gene products.