These studies indicate that a regulatory subset of Iym0hocytes is missing in patients with juvenile rheumatoid arthritis but these patients have antibodies in their serum that react with normal T cells. This regulatory subset of T cells is, however, present in patients whose serum shows little or no reactivity with normal T cells. In addition, patients who are deficient in this regulatory subset of lymphocytes have siguificantly higher numbers of cells secreting Ig as measured by a hemolrtic plaque assay. The significance of these observations old: first, they represent a positive relationship among the loss of regulation, overproduction of immunoglobulin, and the presence of anti-T cell antibodies; and second, and perhaps of equal importance, is the indication that serum from patients with autoimmune diseases may give us a readil av'table reagent with which to dissect further functionally distinct subsets of normal T cells in man. Recent progress in murine systems has demonstrated that both the type and the intensity of the cellular and humoral immune responses can be regulated by a complex series of cellular interactions involving distinct subsets of lymphocytes (1)(2)(3). It has been shown that antigen triggers subpopulations of T lymphocytes capable of helping, amplifying, or suppressing other T-cell or B-cell responses (4, 5). The human T cell, like its counterpart in the mouse, expresses in vitro various functions including proliferation. in response to soluble and cell surface antigens, elaboration of mediators, activation by polyclonal mitogens, and cytotoxic activities (6). In man, it is now clear that regulatory T cells exist and can be defined by their functional properties in both normal (7-9) and pathologic states (10, 11). More recent data support the notion that T-cell help and suppression are mediated by different subpopulations of T cells bearing distinct Fc receptors (12) and cell surface antigens as defined by allo-and heteroantisera (13,14).We have previously shown that sera from four patients with juvenile rheumatoid arthritis (JRA) Patients. Records of all children who have attended the arthritis clinic at the Children's Hospital (Boston, MA) were reviewed. Only those children for whom a positive diagnosis of JRA was confirmed were included in the study (15). There were 30 children with this diagnosis. They were between 3 and 19 years of age and included 5 boys and 25 girls. The time interval between the onset of the disease and the time the blood sample was taken varied between a few weeks (4-6 weeks) and 16 years. The mode of onset was systemic in 3, monoarticular in 9, and pauci or polyarticular in 18. The criteria for classification as active disease or remission were based on objective findings by physical examination alone. Thus, children with at least one swollen joint were considered to have active disease, whether or not they had other physical abnormalities such as limitation of motion, signs of synovitis, or joint deformities. Patients in remission had normal results o...