During the last 7 years significant progress has been made in the identification of melanoma-associated antigens recognized by cytotoxic T lymphocytes (CTL). These antigens belong to three main groups: cancer/testis-specific antigens (MAGE, BAGE, GAGE, PRAME and NY-ESO-1), melanocyte differentiation antigens (tyrosinase, Melan-A/MART-1, gp100, TRP-1 and TRP-2), and mutated or aberrantly expressed antigens (MUM-1, CDK4, beta-catenin, gp100-in4, p15 and N-acetylglucosaminyltransferase V). In this review we have summarized the available data concerning the characterization of melanoma-associated antigens, focusing on their immunogenic and protective properties. The development of a strong immune response to differentiation antigens is limited by the existence of tolerance to these "self"-antigens, permitting the involvement of only T cells with low affinity T-cell receptors. Among the melanoma differentiation antigens, only gp100 has been shown to be a tumor regression antigen. The cancer/testis-specific antigens such as MAGE and PRAME should potentially be highly immunogenic antigens. They contain several potential HLA class I binding epitopes and are present only in the testes, which are not accessible to the cells of the immune system owing to the lack of direct contact with the immune cells and the lack of HLA class I expression on the surface of germ cells. But only two patients have been found who responded to these antigens in vivo, indicating their genuinely low immunogenicity. A comparison of the predicted secondary structures of these two groups of antigens (cancer/testis-specific and differentiation antigens) revealed enrichment of long alpha-helical stretches in the cancer/testis-specific antigens. We hypothesize that such highly organized stable structures could, first, reduce denaturation of the protein and, thus, ubiquitinylation as a degradation signal, and, second, diminish the efficiency of the protein unfolding - a necessary step in the proteolytic cleavage by proteasomes. High structural stability could therefore be responsible for the low immunogenicity of these proteins. In this case, modifications decreasing the stability of these proteins might be a means of improving the immune response to these potentially therapeutically useful antigens.