BACKGROUND Clinical outcome of renal transplantation among systemic lupus erythematosus (SLE) patients remains a topic of controversy. Most of the previous reports were based upon small single-centre studies that were not always well-designed. METHODS We conducted the retrospective analysis using data from USRDS and UNOS databases. Patients were divided into five groups based on the cause of end-stage renal disease (ESRD): diabetes mellitus (DM), SLE, glomerulonephritis, hypertension and other causes. Between 1990 and 1999, 2886 renal transplantation recipients with ESRD due to SLE were identified from a total of 92 844 patients. RESULTS The mean follow-up period of this study was 4.7 +/- 2.4 years. While unadjusted analysis using Kaplan-Meier curves demonstrated an association between SLE and improved allograft survival compared with DM, in multivariate analysis the SLE group had worse allograft [hazard ratio (HR) 1.09, P < 0.05] and recipient (HR 1.19, P < 0.05) survival compared with the DM group. Subgroup analysis based on the type of donor showed that SLE patients who received deceased donor allograft had worse allograft and recipient survival (HR 1.14, P = 0.002 and HR 1.30, P = 0.001, respectively) compared with non-SLE deceased donor allograft recipients. Among living allograft recipients, there were no significant differences in either allograft or recipient survival compared with non-SLE recipients. CONCLUSIONS SLE as a cause of ESRD in renal transplant recipients is associated with worse allograft and recipient survival compared with DM; this association is true for the entire population and for the recipients of deceased donor (but not living donor) transplant. Deceased donor allograft recipients have worse outcomes compared with living allograft recipients.