The disease specificity of antibodies to rheumatoid arthritis nuclear antigen (RANA) was examined by comparing anti-RANA titers in sera from 100 patients with rheumatoid arthritis (RA) with sera from 93 healthy controls. Anti-RANA antibodies were found in 86% of the RA sera and 56% of the controls. The higher titers in the RA sera were unrelated to clinical features or to measurements of circulating immune complexes or rheumatoid factors. To study the relationship of these antibodies to previous Epstein-Barr virus (EBV) infection, antibodies to the EB virus capsid antigen (VCA) were examined and found in 94% of the RA sera and 97% of the adult controls. Four of the six RA sera without anti-VCA antibodies had detectable anti-RANA antibodies, so that we might suggest anti-RANA can arise in the absence of EBV infection. From absorption experiments with non-EBV transformed extracts, we inferred that high anti-RANA titers could be due to reactions with non-Epstein-Barr virus related nuclear antigens. These data cast doubt on current speculation about a possible pathogenic role for Epstein-Barr virus in this disease. Rheumatoid arthritis is characterized by a number of immunologic abnormalities that may represent an abnormal response to an infective agent such as a virus (1). Studies (2) of antibodies in children to conventional and well-characterized viral antigens have failed to demonstrate a specific immune response in rheumatoid arthritis (RA) to any particular virus, including Epstein-Barr virus (EBV). Speculation about a possible etiologic role for EBV in RA has been stimulated by the observations of Alspaugh et a1 (3,4) who demonstrated a precipitating antibody system that was found in 6681% of RA sera and 6 4 % of controls. The reacting antigen RANA, demonstrated by immunofluorescence and immunodiffusion, was detectable only in B cell lines containing the EBV genome or normal human lymphocytes infected with EBV in vitro (5). The claim that anti-RANA antibodies are characteristic of RA has been challenged by their detection with more sensitive techniques in up to 75% of normal sera (6), though the same study has provided stronger evidence of a link with EBV infection by showing that anti-RANA occurred only in normal subjects in whom previous EBV infection could be demonstrated by the presence of antibodies to EB virus capsid antigen (VCA). This finding is at variance with reports (7,8) of RA sera in which anti-RANA antibodies have been described in the absence of previous EBV infection.The purpose of this study was to reexamine the original claim that anti-RANA antibodies are characteristic of RA by using a quantitative immunodiffusion assay to compare titers of anti-RANA antibodies in RA and normal sera, and to examine factors that influence the level of the titers. To investigate the role of EBV infection, the prevalence of antibodies to the