A major peanut allergen, Ara h 2, is recognized by serum IgE from > 90% of patients with peanut hypersensitivity. Biochemical characterization of this allergen indicates that it is a glycoprotein of approximately 17.5 kDa. Using N-terminal amino acid sequence data from purified Ara h 2, oligonucleotide primers were synthesized and used to identify a clone (741 bp) from a peanut cDNA library. This clone was capable of encoding a 17.5-kDa protein with homology to the conglutin family of seed storage proteins. The major linear immunoglobulin E (IgE)-binding epitopes of this allergen were mapped using overlapping peptides synthesized on an activated cellulose membrane and pooled serum IgE from 15 peanut-sensitive patients. Ten IgE-binding epitopes were identified, distributed throughout the length of the Ara h 2 protein. Sixty-three percent of the amino acids represented in the epitopes were either polar uncharged or apolar residues. In an effort to determine which, if any, of the 10 epitopes were recognized by the majority of patients with peanut hypersensitivity, each set of 10 peptides was probed individually with serum IgE from 10 different patients. All of the patient sera tested recognized multiple epitopes. Three epitopes (aa27-36, aa57-66, and aa65-74) were recognized by all patients tested. In addition, these three peptides bound more IgE than all the other epitopes combined, indicating that they are the immunodominant epitopes of the Ara h 2 protein. Mutational analysis of the Ara h 2 epitopes indicate that single amino acid changes result in loss of IgE binding. Two epitopes in region aa57-74 contained the amino acid sequence DPYSP that appears to be necessary for IgE binding. These results may allow for the design of improved diagnostic and therapeutic approaches to peanut hypersensitivity.