Leishmaniasis is a neglected disease with an estimated 12 million infected people. The recent completion of the sequencing of the Leishmania major genome has opened opportunities for the identification of targets for vaccine development. We present here the first attempt at identifying novel vaccine candidates by whole genome analysis. We predicted CD8(+) T cell epitopes from the L. major proteome and validated in vivo in mice the immunogenicity of some of the best predicted epitopes. Consensus epitope predictions from 8272 annotated protein sequences with 5-8 different algorithms allowed the identification of 78 class I CD8(+) epitopes. BALB/c mice were immunized with 26 synthetic peptides corresponding to the most likely epitopes. Fourteen (54%) resulted immunogenic, with eight being strong inducers of T cell IFNgamma production. None of the proteins from which the epitopes are derived are differentially expressed, only two may be surface proteins, eight have putative enzymatic, and metabolic activities. These epitopes and proteins represent new antigen candidates for further studies. While pathogen genomes have not yet delivered their full promise in terms of human health applications, our study opens the way for extensive genome mining for antigen identification and vaccine development against Leishmania and other pathogens.