in 6/7 patients. Steroid dosages were reduced in 9/12 (75%) patients, with a significant reduction from a mean (SD) of 20.2 (9.4) mg/day to 11.6 (5.5) mg/day (p = 0.03).The obvious weakness in this study includes the relatively small numbers of patients who completed the trial, and the follow up was too short to examine the efficacy of leflunomide for long term renal preservation. In addition, our patients were a heterogeneous group with different renal histological classes who had been treated previously with multiple immunosuppressive agents for variable lengths of time. For these reasons, the possibility that the observed improvement, was due in part to corticosteroid or other immunosuppressant drugs given previously cannot be excluded in the absence of a control group. Another limitation includes the correlation of the clinical response with renal histological features that were obtained at a mean of 16 months before leflunomide treatment. Furthermore, repeat renal histology was not available after treatment. In conclusion, leflunomide is a safe and probably efficacious treatment in patients with lupus nephritis who do not respond or cannot tolerate conventional treatments.We thank Aventis for providing the leflunomide that was used in this study.