We have previously shown that p3-13 (KTIAY-DEEARR) of the 65-kDa heat shock protein (hsp65) of Mycobacterium tuberculosis and Mycobacterium leprae is selected as an important T cell epitope in HLA-DR17+ individuals, by selectively binding to (a pocket in) DR17 molecules, the major subset of the DR3 specificity. We have now further studied the interaction between p3-13, HLA-DR17 and four different TCR (V beta 5.1, V beta 1, and V beta 4) by using T cell stimulation assays, direct peptide-DR binding assays, and a large panel (n = 240) of single amino acid substitution analogs of p3-13. We find that residues 5(I) and 8(D) of p3-13 are important DR17 binding residues, whereas the residues that interact with the TCR vary slightly for each DR17-restricted clone. By using N- and C-terminal truncated derivatives of p2-20 we defined the minimal peptide length for both HLA-DR17 binding and T cell activation: the minimal peptide that bound to DR17 was seven amino acids long whereas the minimal peptide that activated T cell proliferation was eight amino acids in length. Furthermore, two new DR17-restricted epitopes were identified on hsp70 and hsp18 of M. leprae. Alignment of the critical DR17-binding residues 5(I) and 8(D) of p3-13 with these two novel epitopes and two other DR17-binding peptides revealed the presence of highly conserved amino acids at positions n and n + 3 with I, L, and V at position n and D and E at position n + 3. D and E are particularly likely to interact with the DR17-specific, positively charged pocket that we have defined earlier. Based on these results, a set of single amino acid substituted analogs that failed to activate these T cell clones but still bound specifically to DR17 was defined and tested for their ability to inhibit T cell activation by p3-13 or other DR17-restricted epitopes. Those peptides were able to inhibit the response to p3-13 as well as other DR17-restricted mycobacterial epitopes in an allele-specific manner, and are anticipated to be of potential use for immunotherapeutic and vaccine design strategies.