Abstract The main conclusion from these experiments is that the antigen‐specific suppressor T cell of mice which inhibits the induction of cytotoxic T lymphocytes is not itself a cytotoxic T cell. This conclusion is supported by two main observations: first, a certain cell number from first‐step cultures which was suppressive in the presence of a high dose of antigen actually helped the cytotoxic response at a lower antigen dose. This observation is difficult to reconcile with the hypothesis that suppression is due to the killing of the stimulator or the responder cells in the second‐step culture by cytotoxic T cells. Second, cells from first‐step cultures of cortisone‐treated mice displayed cytotoxic activity but had no suppressive effect on the generation of killer cells. It was further demonstrated that these cells failed to influence in any way the suppressive effect, however weak, of cells from first‐step cultures of normal spleen. We therefore favor the view that the suppression observed in this system is due to a regulatory signal which occurs as a result of the ability of both inhibitory cells and responder cells to recognize and respond to allogeneic determinants expressed on the surface of stimulator cells. The suppressor T cells described here act by linked associative recognition of antigen. That is, suppressor T cells only inhibit the induction of a precursor cytotoxic T cell in the presence of an antigen to which both the precursor cell and the suppressor cell can bind. In this sense, suppressors act in a manner analogous to helper T cells in T‐B cell cooperation; carrier‐specific helper T cells only enhance an anti‐hapten B cell response in the presence of hapten‐carrier conjugates. Similarly, alloantigen a (carrier)‐specific suppressor T cells only inhibit alloantigen b (hapten)‐specific cytotoxic responses in the presence of (a × b)F1 stimulator cells (hapten‐carrier conjugate), not in the presence of a mixture of parental stimulator cells (a + b).