Infliximab, a chimeric anti-tumor necrosis factor alpha (TNF-alpha) monoclonal antibody, has been recognized as significantly improving the course of rheumatoid arthritis (RA); however, a subset of patients shows poor responses. To understand the mechanism underlying such unresponsiveness, I examined the clinical pharmacokinetics (PK) of infliximab, using time-serum concentration profiles obtained from 21 RA patients who had received infliximab therapy in combination with methotrexate (MTX). At week 14 of therapy, 15 cases achieved good or moderate responses in the European League Against Rheumatism (EULAR) criteria, and 3 cases resulted in nonresponders. The others discontinued therapy because of severe adverse effects or aggravation of disease activities. The means of distribution volume and elimination half-life (t (1/2)) during the first 2 weeks were 0.05 l/kg and 9.5 days, respectively. Through 14 weeks, most good and moderate responders maintained serum concentrations of more than 1 microg/ml, even immediately before the next infusions. Only 3 cases among good or moderate responders showed undetectable levels of trough serum concentration at week 14. In contrast, the PK profiles of all nonresponders except one showed rapid clearance during therapy. These data support the idea that the rapid clearance of infliximab is the main cause of poor therapeutic responses. I also found that the t (1/2) during the first 2 weeks is inversely correlated to the disease activity scores for 28 joints at the start of treatment, suggesting that TNF-alpha levels may determine the disease activity of RA. For patients who showed a rapid clearance of infliximab, the increased use of prednisone or MTX was beneficial to achieve sufficient clinical responses. The addition of tacrolimus was effective to improve the clinical outcomes of nonresponders. Thus PK data apparently offer guidance when modified treatment for infliximab-resistant RA patients is being considered.