The mitogen-induced cell-mediated cytotoxic (MICC) reaction was evaluated in terms of its capacity to identify cytotoxic cells among normal human blood leucocytes. Three conventional mitogens were used (phytohaemagglutinin, pokeweed and concanavalin A) and nine different erythrocyte target cells (chicken, horse, human, ox, rabbit, sheep, mouse, rat and guinea-pig). The effector cells investigated were mononuclear cells, monocyte-depleted mononuclear cells and polymorphonuclear leucocytes. Phytohaemagglutinin consistently mediated the lysis of only chicken erythrocytes whereas pokeweed consistently mediated the lysis of rabbit erythrocytes and, to a much lesser degree, chicken and rat erythrocytes. Con A consistently mediated the lysis of chicken, horse, rabbit, guinea-pig and sheep erythrocytes. Irrespective of the target cell used, PHA and Con A facilitated target cell lysis by monocytes and polymorphs and not by lymphocytes which displayed no cytotoxic activity in the presence of these two mitogens. On the other hand, pokeweed-mediated cytolysis of rabbit erythrocyte target cells was induced by lymphocytes in addition to monocytes and polymorphs. Only monocytes were able to lyse chicken erythrocytes in the presence of pokeweed; polymorphs and lymphocytes were inactive. Agglutination or aggregation of the target cells by the mitogen is not a mandatory concomitant of the MICC reaction. Some RBC may be agglutinated by the mitogen without being lysed by effector cells in the presence of the mitogen. The mechanism of mitogen-mediated cytolysis of target cells remains to be elucidated. These results suggest that the MICC assay can be utilized as a probe to permit the selective detection of functionally distinct subpopulations of cytotoxic monocytes, lymphocytes and polymorphs. Whether these cytotoxic cells are identical to, overlap or are distinct from those cells which mediate the NOCC and ADCC cytotoxic reactions remains to be determined.