Monoclonal antibodies (MoAb's) make it possible to treat rheumatoid arthritis with selective immunotherapy. These antibodies may be directed against various targets, such as lymphocyte activation antigens, cytokines or subpopulations of lymphocytes (notably TCD4 +), involved in the pathogenesis of the disease. Recent open studies have demonstrated the feasibility and safety of this therapeutic method, but the number of patients who entered the trials is still low, and the clinical, biological and immunological results vary considerably in importance and duration, without remission. No response predictive factor could be elicited from these studies. The murine origin of these MoAb's exposes to the frequent risk of immunization which may interfere with the effectiveness and safety of a second treatment. Some possibilities can already be envisaged, including potentiation of the MoAb by coupling with a cytotoxic agent (anti-CD5 + ricin) and "humanization" of murine MoAb's (chimeric anti-CD4) reducing the risk of immunization. Further (controlled) trials therefore are indispensable to evaluate the true rank occupied by this therapeutic method in rheumatoid arthritis.