OBJECTIVE: A number of disease-modifying anti-rheumatic drugs (DMARDs) have been shown to be more effective than placebo in the management of rheumatoid arthritis (RA). However, most course of DMARDs, except methotrexate, are discontinued after 2 or 3 years, because of toxicity, lack of efficacy or escape from control. The multi-drug resistance (MDR) is a phenomenon in which cells develop cross-resistance to many agents such as anthracyclin, vinca alkaloids and colchicine. In our hypothesis, MDR phenomenon could be implicated in acquired resistance to DMARDs in RA. We have established a mdr1 cell line and tested whether DMARDs are substrate for P-glycoprotein (P-gp). METHODS: The mdr1-cDNA was cloned into retroviral vector, and the recombinant retroviral vector was transfected into PA317 cells. The target cells, NIH3T3, were infected with recombinant retroviruses. A colony most resistant to vinblastin was selected for the following experiments; expression of mdr1 gene in NIH3T3 cells was confirmed by RT-PCR, and biological function of mdr1 gene product, P-gp, was tested using Rhodamine-123 (Rh123) efflux assay. Resistance of the target cells expression P-gp which can survive against hydroxychloroquine (HCQ) and methotrxate (MTX) were measured by MTT assay. RESULTS: RT-PCR for mdr1 gene showed successful transfer of the gene into the NIH3T3 cells. Rh123 assay revealed expression of P-gp on the selected cells as follows; Rh123 efflux activity of uninfected NIH3T3 cells was 6%, that of PLXSN was 0.2%, and that of selected cells was 44%. The 50% proliferation inhibitory capacity of the selected cells were twice for HCQ but there was no difference of that for MTX. CONCLUSION: We established a mdr1 cell line and using the cell line, HCQ was a substrate of MDR, but MTX was not related to MDR.