Rheumatoid arthritis (RA) is an autoimmune disease mediated by T-lymphocytes and associated with the human leukocyte antigen-death receptor 4 (HLA-DR4). The HLA-DR4 protein selectively interacts with the antigenic peptides on the cell surface and presents them to the T cell receptor (TCR) on CD4+ T cells. The HLA-DR4-antigen-TCR complex initiates the autoimmune response and eventually causes the chronic inflammation within patients bodies. To inhibit HLA-DR4-restricted T cell activation, an ideal approach is to discover non-T cell stimulating substrates that specifically bind to HLA-DR4. In this paper, a comprehensive structure-based design strategy involved de novo design approach, pharmacophore search, and dock method was presented and applied to "simplify" the known binding peptide ligand of HLA-DR4 and identified specific small-molecule inhibitors for HLA-DR4. The designed three-step strategy successfully identified five nonpeptide ligands with novel scaffolds from a chemical library containing 4 × 10(6) commercially available compounds within a tolerable computing time. The identified five chemicals, BAS-0219606, T0506-2494, 6436645, 3S-71981, and KM 11073, are all non-T cell stimulators and are able to significantly inhibit HLA-DR4-restricted T cell activation induced by type II collagen (CII) 263-272 peptide. IC(50) for the best two potentials, BAS-0219606 and T0506-2494, was 31 and 17 μM, respectively, which is equivalent or better than the known peptide ligands. It is hopeful that they can be used as effective therapeutic means for further treatment of RA patients. In addition, the comprehensive strategy presented in this paper exhibited itself to be an effective flow line from peptide ligands to small-molecule inhibitors and will have applications to other targets.