The aim of the study was: (1) to evaluate the efficacy of the treatment with cyclosporine combined with another disease-modifying drug in RA patients who failed to respond to mono- or polytherapy; (2) to adjust an effective dose of the drug in this treatment; (3) to assess the frequency and nature of adverse effects; and (4) to find out why the therapy had been discontinued. The study comprised 29 patients suffering from rheumatoid arthritis diagnosed according to the ACR criteria, in whom an active inflammatory process continued despite the fact that for 6 months they had been treated with full doses of at least one basic drug. Subjects who had been treated with cyclosporine, with uncontrollable arterial hypertension and kidney insufficiency were not included in the study. At the time of inclusion all the subjects were given steroids at up to 15 mg of prednisolone, and one to three basic drugs. While the therapy with the basic drug was continued, cyclosporine was added in increasing doses starting from 1.5-2.0 mg/kg b.w./24h until a maximum dose of 5 mg/kg b.w./24h was reached. The disease activity was assessed according to modified ACR criteria. An improvement was observed in 24 (82.8%) patients, including 3 (10.3%) in whom the improvement was substantial, 8 (27.6%) where it was intermediate, and 13 (44.8%) – minor. Five (17.2%) patients failed to improve. The dose which was considered effective was 100-300 mg daily (mean 155 mg/24h; i.e. 2.5-3.0 mg/kg b.w./24h). The treatment was discontinued in 14 (48.2%) patients, most often (24.1%) due to adverse events. In 5 (17.2%) patients the treatment was discontinued because of lack of early improvement or delayed aggravation. Cyclosporine used in combination has been proved to be able to improve the condition of patients who do not respond to the therapy with other basic drugs. An effective dose of the drug is about 2.5-3.0 mg/kg b.w./24h. Adverse events leading to the discontinuation of the treatment were observed in every fourth patient, and along with lack of improvement were the most frequent cause of discontinuing the treatment. These results, though, have been biased by the fact that the study comprised subjects in a severe inflammatory state resistant to other forms of therapy. Once this context is considered, this mode of treatment is worth recommending to be applied more widely.