Studies on the TCR repertoire of T cells in several inflamed lesions of SS patients have shown that there exist no unique TCR V family genes, although TCR V gene usage is relatively restricted. Analysis of the clonality of infiltrating T cells shows that some T cells expand clonally in lesions, suggesting that these cells were induced by antigen-driven rather than superantigen-driven stimulation. The restricted usage of the TCRJ beta, V alpha and J alpha genes also supports the notion that all expanded T cells do not accumulate due to superantigen. Junctional sequence analysis has revealed some conserved amino acid sequence motifs in the TCR CDR3 region, which is the antigen-binding site on T cells. These observations strongly suggest that pathogenic T cells in lesions recognize limited epitopes on autoantigens in the context of HLA. These findings are similar to TCR on T cells in rheumatoid synovium, as described elsewhere [51]. Intriguingly, T cells in lacrimal and labial salivary glands recognize common epitopes, but T cells in the kidney react with different autoantigens from those in minor salivary glands. In the future, the peptides recognized by pathogenic autoreactive T cells should be clarified to elucidate the mechanism for the progression of SS. Moreover, we predict that a vaccination with analogue peptides, to induce autoreactive T cells to anergy, might provide a new strategy for the specific treatment of SS.