Lymphocytes were purified from peripheral blood of normal donors and patients with chronic lymphocytic leukemia (CLL) by Ficoll-Hypaque centrifugation. Adenylate cyclase activity, expressed as picomoles [(32)P]cyclic AMP generated per milligram protein per minute, was 57+/-4 in normals and 26+/-4 in CLL patients. Enzyme activity, expressed as picomoles [(32)P]cyclic AMP generated per 10(6) lymphocytes per minute, was 2.09+/-0.19 for normal lymphocytes and 1.10+/-0.16 for CLL lymphocytes. The differences between normal and CLL peripheral lymphocytes are highly significant (P < 0.001) with either method of calculating activity. Cyclic AMP levels (picomoles per 10(6) lymphocytes) also differed significantly: 1.38+/-0.29 for normals and 0.45+/-0.08 for CLL lymphocytes. Adenylate cyclase was assayed in lymphocytes enriched for bone marrow-derived (B) cells by removing E-rosetted thymus-derived (T) cells, and enriched for T cells by harvesting E-rosetted lymphocytes or by removing B cells with nylon wool absorption. Solutions to simultaneous equations gave the following calculated enzyme activities for pure B- and T-cell subpopulations (in picomoles [(32)P]cyclic AMP generated per milligram mg protein per minute): normal B, 196+/-22; normal T, 30+/-10; CLL B, 34+/-6; CLL T, 19+/-4. Thus. normal B-lymphocyte adenylate cyclase exceeds normal T-lymphocyte activity by more than sixfold, whereas in the case of CLL the enzyme activity in B lymphocytes is markedly reduced to levels comparable to T lymphocytes. The responses of lymphocytes to stimulation with the hormones prostaglandin E(1) and isoproterenol, and with NaF, were assessed. Compared with normal lymphocytes, enzyme activities were reduced in CLL lymphocytes incubated with these agents, but to a degree paralleling the reduced basal activities. Thus, the ratios between stimulated and basal adenylate cyclase levels in Ficoll-Hypaque-purified, normal lymphocytes were 2.3+/-0.1 after incubation with 10 muM isoproterenol, and 3.9+/-0.2 with 10 mM NaF, values which did not differ significantly from those obtained with CLL lymphocytes. When the enzyme activities calculated for purified T- and B-lymphocyte subpopulations were used to derive the stimulation ratios, the responses of normal and CLL T and B cells to these agents were also indistinguishable. The simplest explanation for these findings is a reduced number of normally responsive enzyme sites on the surface membranes of CLL lymphocytes, although alternative explanations are possible.