HIV-1 infection has clearly been shown to induce a vigorous CTL response in infected people, and this response is present at a time when immune function otherwise is globally impaired. HIV-1-specific CTL are detectable both in peripheral blood and tissues of infected people, and are aimed at multiple viral proteins. The precise epitopes recognized by these CTL are now being defined, and the establishment of CTL clones should facilitate further functional analysis of these cells. However, the central question as to the clinical relevance of HIV-1-specific CTL remains. By analogy with animal model systems of virus infection, it is reasonable to postulate that HIV-1-specific CTL serve a protective role as a host defense. In this regard, in vitro data indicate that HIV-1-specific CTL can suppress viral replication, and longitudinal clinical studies indicate that the vigorous CTL activity seen in the early stages of infection declines with disease progression. Alternatively, the presence of HIV-1-specific CTL in tissues such as the lung and brain have to at least raise the possibility that these cells may be contributing to the pathologic consequences of infection. In addition, the relative protective effects of virus-specific CTL compared to other effector mechanisms such as ADCC and neutralizing antibodies remain to be determined. Nevertheless, recent data in the SIV vaccine model give reason for encouragement that a state of protective immunity can be achieved in AIDS-like illness caused by retroviruses. The search continues presently not only for the parameters which define protective immunity in HIV-1 infection, but also for the ideal HIV-1 immunogens to be used for vaccination of human populations.