Eighteen rheumatoid arthritis (RA) patients, who had been treated with nonsteroidal antiinflammatory drugs (NSAIDs) only, were enrolled in a 12-week, open-label, randomized protocol to determine whether clinical responses might be associated with improvement in laboratory measures of inflammation and immunologic activity in RA patients treated with NSAIDs. Following a 2-week drug washout period, patients were given either long-acting ibuprofen (2,400 mg/day) or flurbiprofen (200 mg/day); clinical, laboratory, and immunologic assessments were done biweekly for 10 weeks. A clinical efficacy index, utilizing a combination of measures of disease activity, identified 7 "responders" and 10 "nonresponders" (1 patient discontinued therapy because of a rash). Flow cytometric analysis revealed no abnormalities in the numbers of circulating CD3+, CD4+, or CD8+ lymphocytes in the 17 patients. The density of these T cell markers at enrollment was similar in patients and control subjects. However, following the 2-week drug washout, significant worsening of morning stiffness, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) was accompanied by a significant decrease (P less than or equal to 0.05) in the density of these T cell surface determinants, as is characteristic of activated T cells. After 10 weeks of NSAID therapy, increased density of CD3, CD4, and CD8 was observed in 47%, 73%, and 50% of the patients, respectively. However, in only the responders was the density of these T cell surface markers increased significantly (P less than or equal to 0.04). The responders, but not the nonresponders, also demonstrated significant reductions in the ESR, CRP level, serum IgM rheumatoid factor (RF) titer, and spontaneous synthesis of RF by lymphocytes in vitro (P less than or equal to 0.05). There were significant correlations between improvements in the clinical parameters (50-foot walk time, joint score, and global assessment) and reductions in the ESR, CRP level, and serum RF titer (P less than or equal to 0.05). These findings demonstrate that clinical improvement in RA patients treated with NSAIDs may be associated with the disappearance of phenotypically activated circulating T cells and functionally activated B cells, as well as with reductions in acute-phase reactants and serum RF titers.