Lymphocytes binding 1251-labeled phytohemagglutinin ( i251-PHA) were directly quantitated by autoradiography. The distribution of grain counts proved t o be trimodal indicating the presence of three distinct subpopulations of PHA-binding cells. Low intensity binders (median grain count of 8-12 grains/cell) predominated in spleen and lymph node, included B and T cells, and responded to the mitogen, lipopolysaccharide (LPS). High intensity binders (median grain count greater than 60 grains/cell) occurred at low frequency in every lymphoid organ, were T cells, and included mixed lymphocyte culture (MLC)-reactive but not PHA-reactive cells as judged by preferential sensitivity to suicide with lz5I-PHA. Medium intensity binders (median grain count of 40-45 grains/cell) predominated in the thymus, formed a subpopulation in the spleen and lymph nodes of normal mice, but were absent from bone marrow of normal mice and the spleen of nude mice. In addition, their numbers were markedly reduced following adult thymectomy and cortisone treatment indicating that they were most likely T cells of recent thymic origin.PHA-reactive cells were considered t o belong to the medium intensity PHAbinding subpopulation. This conclusion was based on the following: first, medium intensity binders, like PHA-reactive cells, were T cells; second, they were contained in the same subpopulation of T cells as PHA-reactive cells. That is, they were of recent thymic origin (T, cells) according to studies of tissue distribution, cortisone sensitivity and effect of adult thymectomy; third, 'suicide' of the response t o PHA as well as to allogeneic cells but not LPS was achieved by prior incubation of cells with high doses of soluble '2SI-PHA, whereas lower doses resulted only in abrogation of the MLC response. In other words, PHA-reactive cells resided in the medium, rather than high intensity subpopulation of binding cells. By analogy with suicide of antigen-binding cells, this observation implies that PHA binds preferentially to those cells responding to it, and confirms the validity of using mitogen activation as a model for antigen-induced triggering. On the other hand, binding of PHA t o a subpopulation of T cells casts doubt on the value of the PHA response as a measure of overall T cell function. Support for this conclusion came from the failure of PHA suicide to abrogate effector T cell activities such as help and suppression.