The short-term outcome of patients with diffuse proliferative lupus nephritis (DPLN) has improved with advances in immunosuppressive treatment. However, the impact of different immunosuppressive regimens on long-term renal function remains to be defined. This prospective cohort study examined the long-term renal function and disease relapse in adults with biopsy-proven DPLN, significant proteinuria, and hypoalbuminemia, who had been treated with sequential immunosuppression comprising prednisolone and oral cyclophosphamide as induction followed by low-dose prednisolone and azathioprine as maintenance. Sixty-six patients with 68 episodes of DPLN were included, with follow-up of 91.7 +/- 36.7 months. 82.4% achieved complete remission and 39.1% relapsed during follow-up. Patients in partial remission were at higher risk of relapse compared with those in complete remission (hazard ratio 6.2, P < 0.001). Serum creatinine remained stable over time (P = 0.931), while creatinine clearance showed a significant increase with time after treatment (P = 0.032). Three (4.4%) patients had doubling of baseline creatinine, but none reached end-stage renal failure or died. Univariate and mixed model analyses showed that the evolution of long-term renal function was significantly influenced by the chronicity score and creatinine clearance at baseline, and by the renal function at one year after treatment. These data demonstrate the efficacy of sequential immunosuppression in preserving renal function in most Chinese subjects with DPLN. The results also indicate that irreversible renal scarring (as reflected by baseline chronicity score), renal reserve (as reflected by renal function at baseline and one year), and an induction regimen that is effective in preserving the nephron mass are critical determinants of long-term renal outcome.