In this study, human derived melanoma cell lines (518.A2, Gerl43 and DAUV) were fused with EBV B-LCL (HMy2) using polyethylene glycol as the fusogen. Although the hybrid cells had a dominant tumour cell phenotype and did not express the co-stimulatory ligand molecules CD80 or CD86, they showed a significant increase in their ability to stimulate primary allogeneic T cell responses in vitro, as compared with the parent melanoma cells. The MHC class I(+), MHC class II(-) hybrid cells (518.A2 x HMy2) were able to directly stimulate separated CD8+ T cells, but not CD4+ T cells, whereas the MHC class I(+), MHC class II(+) hybrid cells (HMy2 x Gerl43) were able to directly stimulate both CD4+ and CD8+ T cell subsets. The T cell response was independent of CD80/CD86 interaction with CD28/CTLA-4, as the response could not be blocked with CTLA-4 Ig. Bystander non-T cell co-operation enhanced the T cell responses to the hybrid cells, by providing CD80/CD86 dependent co-stimulatory signals. Genetic modification of Gerl43 and the HMy2 x Gerl43 hybrid cells to express CD80 significantly increased the level of direct CD3+ T cell proliferation, compared with the non-transfected cells. The HMy2 x 518.A2 and HMy2 x Gerl43 hybrid cells retained expression of tumour-associated antigens, as determined by RT-PCR, and were antigenically stable in vitro. HMy2 x 518.A2 clone 2 expressed tumour-associated antigens at greater levels than the parent melanoma cells, and was able to present MAGE 1 and MAGE 3 antigens to HLA class I-restricted, antigen specific CTL clones with grater efficiency than the parent melanoma cell line. In addition, HMy2 x 518.A2 clone 2 was recognised by two anti-tyrosinase specific CTL clones, despite the expression of this antigen not being detected by RT-PCR. Overall, the data presented in this study supports the idea that APC x tumour cell hybrids have potential as candidates in tumour immunotherapy.