The aims of this study were to define the T-cell subpopulation(s) detected by the virus plaque assay, and particularly to determine whether the virus plaque assay could be used to enumerate cytotoxic T lymphocytes. In addition, studies were undertaken to ascertain whether cell proliferation was required for development of cytotoxic effector function and virus plaque formation by these subpopulations. The results of experiments with a secondary mouse mixed lymphocyte culture (MLC) model indicated that 70 percent of virus plaque-forming cells bore the Ly 1 phenotype and 30 percent the Ly 2,3 phenotype. Three lines of evidence suggested that cytotoxic T lymphocytes (CTL) can be detected by this assay: the fact that some virus plaque-forming cells (V-PFC) bear the same Ly phenotype as CTL; the use of an inhibitor of DNA synthesis indicated that proliferating cells could be eliminated with no effect on V-PFC production and cytotoxic activity of the Ly 2,3 cell population; and that infection of primed lymphocyteswith vesicular stomatitis virus before (MLC) stimulation eliminated cytotoxic activity. In primary MLC, development of V-PFC and CTL was completely abolished by cytosine arabinoside. In contrast, in secondary MLC, some CTL and V- PFC were generated by antigenic stimulation in the absence of proliferation. However, the development of both functions became progressively more susceptible to cytosine arabinoside as the time between primary immunization and in vitro boosting is increased. It is suggested that there may be a considerable disparity between the number of existing effector cells at any given time and the cytotoxic potential, i.e. the number of cells capable of being generated by antigenic stimulation.