The efficacy of the phosphodiesterase (PDE) IV inhibitor rolipram on antigen-induced arthritis (AIA) in mice was evaluated in comparison with clinically used anti-arthritic drugs. To induce AIA, DBA/1 mice were immunized with ovalbumin (OVA) emulsified with CFA (day 0) followed by intra-articular injection of OVA on day 21. Rolipram and clinically used anti-arthritic drugs including indomethacin (IND), dexamethasone (DEX), methotrexate (MTX), auranofin (AUR), and D-penicillamine (D-PA) were orally administered daily from days 0 to 20. On day 22, anti-OVA IgG in serum, proliferative responses of spleen cells to the OVA, and anti-OVA IgG2a and interferon (IFN)-gamma as indicators of Th1 responses, as well as anti-OVA IgG1 and interleukin (IL)-10 as those of Th2 reactions, were measured. Treatment with rolipram was followed by inhibition of the early phase of AIA associated with downregulation of both OVA-specific splenocyte proliferation and decreases of IFN-gamma released from the spleen cells but no decreases of the amount of IL-10, or levels of anti-OVA IgG, IgG2a, and IgG1. All clinically used anti-arthritic drugs were more effective in suppressing the late phase of AIA compared with the early phase of joint inflammation. The suppression of AIA by clinically used anti-arthritic drugs was associated with down-regulation of not only Th1 but also Th2 responses. These results suggest that PDE IV inhibitors such as rolipram may exert their suppressive effects on AIA with relatively selective downregulation of antigen-specific Th1 responses compared with anti-arthritic drugs.