Vaccination with tumour cells genetically modified to support induction of an immune response either by production of cytokines or expression of co-stimulatory molecules provides a promising therapeutic approach. We have evaluated the efficiency of tumour vaccination using RENCA cells, a renal cell carcinoma of the BALB/c strain, which were stably transfected with MHC class II, B7.1 or both. Tumour growth after vaccination with MHC class II and/or B7.1 transfected RENCA cells was extremely variable, with protection close to 100% after vaccination with some clones and no effect of vaccination with others. To unravel the underlying mechanism, untransfected RENCA cells were cloned, and individual clones were tested for immunogenicity; that cloned RENCA cells varied considerably in immunogenicity. Whereas all clones displayed comparable growth rates in nude mice, some grew very slowly in immunocompetent syngenetic hosts. Vaccination with rapidly growing clones was ineffective and, importantly, this feature remained unaltered by vaccination with MHC class II and/or B7.1 transfected clones. Instead, 8 of 10 mice rejected the parental line after immunisation with a pool of MHC class II and B7.1 transfected clones. Finally, by cloning RENCA cells, we obtained one highly immunogenic clone (P2). Vaccination with this clone led to an individual-specific response, which indicates that during the cloning procedure a new strongly immunogenic entity must have arisen. Taken together, our results indicate that vaccination with MHC II and/or B7.1 transfected tumour cells induces an efficient immune response, but only if the tumour is weakly immunogenic. Since tumours may be composed of clones displaying different antigenicities, it is mandatory to use bulk cell populations for transfection and vaccination.