The immune system responds efficiently to bacteria, viruses and other agents however, the immune response to cancers is not as effective. In most cases other than specific genetic rearrangements leading to nonself proteins such as in leukemia and idiotypes in lymphoma, tumor associated proteins are self proteins and are not recognized by the immune system to prevent malignancy. In most cancers, patients develop antibodies and/or CTL-precursors to tumor associated antigens but are not effective in generating a therapeutic immune response. Adjuvants have been used with either whole tumors, subunits or peptides with the aim of increasing their immunity. Whole tumor antigens have certain advantages associated with it, such as ready availability as recombinant proteins, potential epitopes that can be presented by a number of MHC class I/II alleles and antibody development. The methods of identification of CD8 and CD4 epitopes either by use of epitope prediction algorithms or use of transgenic mice has made the use of defined synthetic peptides more attractive. The possibility to synthesize long peptides and introduce multiple epitopes (CD4 or CD8) from single or multiple antigens makes peptide a viable alternative to whole proteins. As an alternative to totally synthetic peptide constructs or polymers, polytopes have been generated by genetic engineering methods. In addition, to deliver immunogens to and to activate DC, receptor-mediated delivery of peptides using antibodies, cytokines and carbohydrates have been used. This review will encompass the various strategies, preclinical and clinical applications in designing peptide-based vaccines for cancer.