The contention that cancers are immunogenic is still being debated: one view states that cancers evolve to evade immune surveillance, while the other claims that cancers actively induce tolerance. A unifying hypothesis is that immune surveillance operates early on in carcinogenesis, but eventually tumours adapt to shift the balance from activation to tolerance. Cancer vaccines therefore, must stimulate a tumour specific response and break active tolerance mechanisms. Cancer vaccines were initially comprised of tumour associated antigens (TAAs), but since these were poorly immunogenic, they were replaced with dendritic cell (DC) vaccines loaded with TAAs (adopted because DC prime naive T cells). However, such DC-based vaccines are impractical to produce and have not proved particularly effective. These disappointing results have led to the search for more innovative vaccine strategies. Viral vectors encoding TAAs are promising vaccine candidates because they can infect antigen presenting cells (APCs) in vivo they may also signal "danger" to the immune system through pathogen recognition pathways. This thesis is concerned with the development of vaccines based on lentiviral vectors (lentivectors) such vectors are readily used for gene therapy because they can stable modify non-dividing cells. The aim of this project was to assess their ability to deliver antigens to dendritic cells in vivo and to stimulate effective T cell responses. The findings of this thesis are that: Directly injected lentivectors can stimulate CD4+ (and CD8+) T cell responses to their encoded antigen if it is targeted into the endocytic pathway. DC stably present antigen in vivo following vector immunisation. DC can be induced to mature by including an NF-kB activator (vFLIP) in the vector. These results support the development of lentivirus-based cancer vaccines.