While T cells have been clearly implicated in a number of disease processes including autoimmunity, graft rejection, and atypical immune responses, the precise Ags recognized by the pathogenic T cells have often been difficult to identify. This has particularly been true for MHC class II-restricted CD4+ T cells. Although such cells can be demonstrated to have undergone clonal expansion at sites of pathology, they are frequently difficult to establish as stable T cell clones. Furthermore, in general, larger peptides in higher concentrations are required to stimulate CD4+ T cells than CD8+ T cells, which makes some of the techniques developed to identify CD8+ T cell Ags impractical. To circumvent some of these problems, we developed a model system consisting of two parts. The first part involves the construction of an indicator T cell hybridoma expressing a chimeric TCR comprised of murine constant regions and human variable regions specific for influenza hemagglutinin 307–319 presented by DR4. The second part consists of a library of fibroblasts each expressing multiple peptides as amino terminal covalent extensions of the β-chain of HLA-DR4 (DRA1*0101, DRB1*0401). Using this model system, we screened ∼100,000 peptides and identified three novel peptides stimulatory for the HA1.7 TCR. While there is some convergence at residues known to be important for T cell recognition, all three peptides differ markedly from each other and bear little resemblance to wild-type hemagglutinin 307–319.