To learn whether rheumatoid factor (RF), HLA-DR4, or current therapy influences clonal expansion of B lymphocytes (B cells) in persons with rheumatoid arthritis (RA).We measured clonal expansion by analysis of cell surface staining for immunoglobulin light chains. Double staining methods detected a B cell marker (CD19) plus either kappa or lambda on peripheral blood lymphocytes from subjects with RA (n = 26) and controls (n = 26). The difference between frequency histograms of surface kappa and lambda staining was determined by the Kolmogorov-Smirnov statistic D that represents the fraction of clonally expanded B cells.The mean D value in RA was over 50% higher than in the controls [0.225 +/- SD 0.155 versus 0.144 +/- 0.025 (p < 0.0001)]. Ten subjects with RA had values exceeding +2 SD for controls (p = 0.0007). Mean D correlated with RF titer (Spearman's rank correlation coefficient rSp + 0.53, p = 0.006). All 10 high D values were found in the RA subgroups with positive serum tests for RF and with the HLA-DR4 positive genotype. The channel of maximal difference between kappa and lambda staining was higher in the RA group than in controls, showing that clonal expansion was most marked among brightly staining cells. Patients with RA currently receiving low dose methotrexate (MTX) tended to have higher D values than those not receiving MTX (mean 0.29 versus 0.18, respectively, p < 0.025). The RA group currently receiving MTX had a higher frequency of abnormal D values (7 of 11 versus 3 of 15 not currently receiving MTX, p = 0.03). This probably reflects preferential use of MTX for severely affected individuals. Confirmatory studies to detect clonal immunoglobulin gene rearrangements were attempted in selected individuals with high D values, but none was demonstrated in total leukocytic or B cell enriched fractions.Findings consistent with B cell clonal expansion occur in about 40% of persons with RA, particularly in the subgroups with positive serum tests for RF and with the HLA-DR4 genotype. However, the clonal expansion level must be below the sensitivity of confirmatory methods.