Peptide binding to major histocompatibility complex (MHC) molecules is a prerequisite for initiating an immune response. This article first describes an approach to predict MHC-peptide-binding sites by using an evolutionary algorithm (EA). The predicted binders are subsequently characterized for their physicochemical properties. The details and implementation issues of the peptide-binding prediction technique are discussed, and the performance comparison with the existing methods is provided. The binding motif derived in silico is used to characterize the physicochemical properties of the experimentally determined binders.