We investigated the antigenic requirements for restimulation of H-2- restricted cytolytic T lymphocytes (CTL) in vitro to determine whether H-2 I region-restricted helper T cells are required in these responses. In one set of experiments, we studied the in vitro response of (responder x nonresponder)F(1) female T cells to the male antigen H-Y. We chose to examine this response because it has been suggested that the defect in nonresponder strains is a failure of helper T cells to recognize H-Y in association with nonresponder I region determinants. However, we find that nonresponder male stimulator cells are as effective as F(1) male stimulator cells at inducing H-Y-specific CTL responses. This finding calls into question reports that secondary CTL responses to H-Y are dependent upon the activation of H-Y- specific helper T cells restricted to responder type I region determinants. In a second set of experiments, we examined the requirements for restimulation of H-2-restricted T cells specific for minor-histocompatibility antigens from long-term mixed lymphocyte cultures. These cultures were established by repeatedly restimulating cultures of specific T cells with H- 2-matched stimulator cells expressing foreign minor histocompatibility antigens. We found that H-2D-restricted T ceils, including CTL, could be restimulated with cells that were matched with the responding cells at only the D region genes. This response did not appear to result from positive allogeneic effects or from antigen processing and "representation" by responder type APC that might contaminate the cultures. Thus, we find no evidence for a requirement for I region-restricted helper T cells in these CTL responses. However, helper T cells are required because we find that CTL lines derived by limit-dilution cloning from these long-term MLC are absolutely dependent upon exogenous helper factors for growth. The most simple interpretation of these results is that the helper cells are restricted to H-2 antigens other than I region antigens or to antigens that code outside of the H-2 complex. Finally, we show that factor-dependent CTL lines must recognize their specific antigen to proliferate, even in the presence of exogenous factors. The requirement of activated CTL for antigen to proliferate provides an explanation for how specific CTL can be selectively enriched in MLC by specific antigen stimulation. Furthermore, it is at variance with reports that memory CTL or activated CTL require only interleukin 2 for restimulation.