Although the efficacy of cyclosporine therapy in rheumatoid arthritis has been established, there have been no long term studies of the risk/benefit ratio of cyclosporine A in severe rheumatoid arthritis. A prospective, open-label one-year study included 106 patients (83 women and 23 men; mean age 53 years; mean disease duration, 11 years) with rheumatoid arthritis. Mean number of previous second-line treatments was four and 69% of patients had failed methotrexate therapy. The initial dosage of cyclosporine was 3 mg/kg/d and was increased if needed up to 5 mg/kg/d. The dosage was reduced in the event of serum creatinine elevation (by more than 30% versus baseline) or diastolic blood pressure elevation (above 95 mmHg). The statistical analysis was performed on an intention-to-treat basis. In the 45 patients who completed the one-year study period, the mean dosage was 3.6 +/- 1 mg after six months and 3.3 +/- 1 mg/kg/d after one year. Significant improvements were seen in all the clinical efficacy parameters. The mean reduction in corticosteroid dosage was 0.5 mg/d. The study drug was discontinued prematurely in 61 patients (36 because of adverse events and 21 because of inefficacy). Twelve of the 56 patients with serum creatinine level elevation on at least one occasion and seven of the 35 patients with diastolic blood pressure elevation were taken off the study drug.