Cancer immunotherapy has traditionally undergone a 'revolution' every decade, from the use of Bacille Calmette-Guérin by scarification in the 1970s, to interleukin-2 therapies in the 1980s, and monoclonal antibody treatments in the early 1990s. Usually the early reports on the use of such agents were encouraging, but when more patients were studied in multiple centres, the initial promising results could not be confirmed. Now in a new century, we have more reagents and methods available than ever before - indeed, with such a plethora of reagents it is difficult to envisage them being fully and appropriately tested within the next decade, by which time there will be even more reagents to test. However, there have been three major advances which should lead to substantial progress in cancer immunotherapy: (1) the widespread use of genetic engineering, enabling identification of candidate vaccine proteins and manipulation of their sequences; (2) the production of antigens, antibodies and cytokines in large amounts by recombinant technologies, and (3) an understanding of the mode of presentation of peptides by major histocompatibility complex Class I and Class II molecules and their recognition by T cells. Despite these advances, there are major problems facing cancer immunotherapy, such as the ability of tumours to mutate and evade the immune system and the difficulty of precisely defining the interactions of effector cells in mediating 'rejection' or destruction of a tumour. There are clearly immunological similarities with diseases such as malaria and schistosomiasis, where the invading foreign organisms can use a variety of strategies to resist an elicited immune response. The failure to find a suitable vaccine for these diseases must lead to some pessimism for the development of immunotherapy for an autologous tumour. However, there are promising studies now in progress which should give an indication of the most important directions to follow. This review provides a commentary on aspects of cancer immunotherapy and in particular will deal with: (1) the selection of antigens as vaccine components; (2) the modes of presentation of antigens, particularly by major histocompatibility complex Class I molecules; and (3) new modes of delivery of vaccine immunogens.