Several recent open studies suggest that the response rates of lupus nephritis to intravenous (IV) cyclophosphamide are lower than those observed in clinical trials. One explanation could be ethnic differences; for example, black patients more frequently have treatment-resistant lupus nephritis. Another could be the inclusion of patients who are noncompliant with therapy. From our register of 268 systemic lupus erythematosus (SLE) patients examined between 1973 and 1996, 61 patients were treated for proliferative lupus nephritis (17 had World Health Organization [WHO] type III and 43 had WHO type IV) and were followed through to 2001. Exclusion criteria included a serum creatinine level >3 mg/dL. In this retrospective study, we assessed renal outcome and survival with an endpoint of end-stage renal disease (ESRD) or death (Kaplan-Meier). In the univariate analysis, worse prognostic factors for survival were serum creatinine >1.3 mg/dL (p < 0.001), age <30 years (p < 0.001), class 2 renal function stage (p < 0.03), and renal biopsy activity index >7 (p < 0.02). In the subgroup of 26 patients treated with IV cyclophosphamide, survival at 5 and 10 years was 82% and 73%, respectively. The dosage of IV cyclophosphamide was slightly lower than usual and used for a shorter period (median = 23 months) than what is usually recommended because of the high frequency of complications. Renal outcome of the IV cyclophosphamide-treated patients was poorer than that reported in the National Institutes of Health series (ESRD: 15% versus 3%). This low survival rate could reflect the short course and lower doses of IV cyclophosphamide used or ethnic differences. These data emphasize the need for continuous research for better-tolerated drug schemes for treatment of our lupus nephritis patients.