Two-color FACS analysis was used to study activated and “functional” T and natural killer (NK) cell subsets of circulating lymphocytes in 23 patients with B-type chronic lymphocytic leukemia (B-CLL) and in 30 healthy subjects. As compared with controls, B-CLL patients had increased absolute numbers of phenotypically activated, HLA-DR+ CD4+ and CD8+ cells and T suppressor/effector (CD11b+CD8+) cells. When patients in Rai stages II through IV (n = 11) were compared with cases in Rai stages O through I (n = 12), the former group of patients had higher numbers of activated CD4+ and CD8+ T cells and decreased levels of suppressor/effector T cells. The absolute numbers of T suppressor/inducer (CD45R+CD4+) cells were elevated in patients with stage O through I disease but within normal range in stage II through IV leukemia. We further showed that the absolute numbers of NK-like (CD16+) cells and their activated counterparts (DR+CD16+) are elevated in B-CLL patients as compared with healthy subjects. The comparison of relative T and NK subsets in the blood of patients and controls showed that the proportions of CD4+, CD8+, and CD16+ cells expressing the activation marker HLA-DR were increased in B-CLL. Furthermore, the percentage of T-suppressor/inducer (CD45R+) cells within the CD4+ population was decreased in the patients. The proportion of T- suppressor/effector (CD11b+) cells within the CD8+ subset was reduced in subjects with stage II-IV disease only. When stimulated in vitro with the T-cell-dependent inducer TPA, B-CLL cells from patients in Rai stages II through IV secreted larger amounts of IgM as compared with cells from stage O through I patients. A positive correlation was observed between the degree of phenotypic activation of CD4+ T-helper cells and their functional capacity to augment IgM secretion by autologous B-CLL cells. Our findings indicate a tumor cell-directed regulatory role of T cells (and possibly NK cells as well) in B-CLL. Furthermore, monitoring of phenotypically activated and functional T- cell subsets may be helpful in the prediction of disease progression and timing of therapy in B-CLL.