Development of a vaccine that provides sterilizing immunity against HIV infection remains an elusive goal, due primarily to the difficulty in generating neutralizing antibodies to primary HIV isolates. In lieu of a present solution to this problem, recent approaches to develop vaccines against HIV/AIDS have focused not on preventing infection outright, but on eliciting potent antiviral CD8+ T-cell responses to limit HIV replication in individuals who become infected after vaccination. Successful control of HIV replication in vivo, enabled by vaccine-elicited immune responses should, in turn, attenuate an individual's rate of progression to AIDS while reducing their likelihood of subsequently transmitting HIV. Recent pre-clinical evaluation of CTL-based vaccines in non-human primate models of AIDS has shown several different vaccine modalities (e.g. heterologous 'prime/boost' strategies such as DNA + recombinant viral vectors) to be capable of eliciting high-level cellular immune responses that are associated with limitation of virus replication and protection against disease following challenge with select pathogenic virus isolates. However, it is not currently known to what extent these protective effects, observed under optimal experimental conditions in select animal models, can be translated into relevant protection of humans against AIDS. In this article we discuss the promise, potential limitations, and scientific challenges that currently provide the context for efforts to develop and successfully employ a safe and effective AIDS vaccine.