The role of the thymus in the cyclical appearance of the dominant idiotype of the myeloma protein secreted by the TEPC-15 plasmacytoma (T-15)-bearing plaque-forming cells (PFC) and anti-idiotypic cells (i.e., cells with receptors for T-15) in the spleen during a primary response to the phosphorylcholine determinant of Streptococcus pneumoniae, strain R36a (Pn) was studied using normal mice, thymus-deficient nude mice, and thymus gland-grafted nude mice (TG-nude). The nude mice and their phenotypically normal littermates (LM) were backcrossed on the BALB/c genetic background. The kinetics of the anti-Pn PFC response of BALB/c inbred mice, littermates of nude mice, and TG-nude mice were essentially the same. There was an initial peak on day 5-6 followed by a decline to near background, and then a second peak on day 12. In nude mice, the first peak of anti-Pn PFC (day 5) was comparable in magnitude to that of mice with an intact thymus; however, there was no second peak. In contrast to the cellular response measured at the level of PFC, the serum antibody response to Pn (assayed by passive hemagglutination of sheep erythrocytes coated with Pn polysaccharide) was comparable in all groups of mice and did not show a measurable oscillation. The anti-idiotypic cellular activity was determined by the ability of spleen cells to bind radiolabeled (125I) TEPC-15 myeloma protein (IgA, kappa) which carries an idiotypic determinant indistinguishable from that of most anti-phosphorylcholine antibodies in BALB/c mice. Binding of radiolabeled McPC-603 (IgA, kappa) and MOPC-315 (IgA, lambda 2) myeloma proteins (which lack the T-15 idiotypic determinant) served as controls. The changes in T-15 binding by splenic lymphocytes following the Pn immunization differed between normal and athymic mice. BALB/c, LM, and TG-nude mice showed a biphasic pattern with peaks at days 3--4 and 10--11 that was nearly reciprocal to the PFC curve. On the other hand, T-15 binding in nude mice either declined and remained depressed or was not affected by the ongoing anti-Pn response. These observations demonstrate that mature T cells are required for cyclical idiotypic and anti-idiotypic responses to immunization with a T-independent antigen and suggest that the cyclical immune response may result from an interaction between idiotypic and anti-idiotypic cell clones.