The emergence of immunotherapy as an important tool in the fight against cancer takes advantage of the exquisite selectivity of antibodies. Targets, however, are limited to those on the cell surface, while most driver mutations occur in the genes encoding intracellular proteins. To overcome this limitation, antibodies can be engineered to target mutation derived neoantigens, peptides derived from mutant proteins that are presented on the cell surface by the Major Histocompatibility Complex Class I (MHC-I). In cases where both the wild-type and mutant peptide are presented, antibodies can be developed that selectively target the mutant peptide-MHC-I complex. We have developed scFvs that target peptides derived from the S45F mutation in β-catenin and the R140Q mutation in IDH2. We determined structures of both the wild-type and mutant peptide-bound MHC-I complexes and discovered that while the phenylalanine residue in the β-catenin peptide is exposed and available for antibody binding, the glutamine residue in the IDH2 peptide is buried within the MHC-I complex, and binds in an almost identical conformation to the wild-type peptide. Determination of the structure of the Fab/peptide-bound MHC-I complex revealed a subtle, induced backbone movement in the mutant peptide that contributes to the observed antibody selectivity. Future work is concentrating on exploiting our detailed structural understanding of the mechanisms of selectivity to develop more selective and potent antibodies.