Whole tumour cells are a logical basis for generating immunity against the cancers they comprise or represent. A number of human trials have been initiated using cytokine-transfected whole tumour cells of autologous (patient-derived) or allogeneic [major histocompatibility complex (MHC)-disparate] origin as vaccines. Although precedent exists for the efficacy of autologous-transfected cell vaccines in animal models, little preclinical evidence confirms that these findings will extrapolate to allogeneic-transfected cell vaccines. In order to address this issue a murine melanoma cell line (K1735) was transfected to secrete interleukin (IL)-2, IL-4, IL-7 or granulocyte-macrophage colony-stimulating factor (GM-CSF); cytokines currently in use in trials. The efficacy of these cells as irradiated vaccines was tested head-to-head in syngeneic (C3H) mice and in MHC-disparate (C57BL/6) mice, the former being subsequently challenged with K1735 cells and the latter with naturally cross-reactive B16-F10 melanoma cells. Whilst the GM-CSF-secreting vaccine was the most effective at generating protection in C3H mice, little enhancement in protection above the wild-type vaccine was seen with any of the transfections for the allogeneic vaccines, even though the wild-type vaccine was more effective than the autologous B16-F10 vaccine. Anti-tumour cytotoxic T-lymphocyte (CTL) activity was detected in both models but did not correlate well with protection, whilst in vitro anti-tumour interferon-gamma (IFN-gamma) secretion tended to be higher following the GM-CSF-secreting vaccine. Cytokine transfection of vaccines generally increased anti-tumour CTL activity and IFN-gamma secretion (T helper type 1 response). Further studies in other model systems are required to confirm this apparent lack of benefit of cytokine transduction over wild-type allogeneic vaccines, and to determine which in vitro assays will correlate best with protection in vivo.