Objectives The aim of this study was to study the relationship between immunosuppressive drug treatment and survival in patients with systemic lupus erythematosus (SLE). Methods Patients who fulfilled four or more American College of Rheumatology criteria for SLE were followed longitudinally. Clinical characteristics, use of immunosuppressive agents and mortality were reviewed. Cox regression was used to study the relationship between immunosuppressive treatment and survival, adjusted for age, sex, vascular risk factors, organ damage, the anti-phospholipid antibodies and a propensity score for the indication of individual immunosuppressive agent derived from separate regression models. Results A total of 803 SLE patients were studied (92% women; age of SLE onset 33.2±14 years; follow-up time 10.8±7.7 years). The frequencies of ever use of immunosuppressive agents were: high-dose prednisolone (≥0.6 mg/kg/day for ≥4 weeks) (85%), azathioprine (63%), cyclophosphamide (25%), mycophenolate mofetil (27%), the calcineurin inhibitors (23%) and hydroxychloroquine (69%). Ninety-seven patients (12%) died and 56 (7%) patients were lost to follow-up. The causes of death were infection (44%), cerebrovascular events (12%), cardiovascular events (10%) and malignancy (8.2%). Cox regression revealed that the ever use of high-dose prednisolone, mycophenolate mofetil, calcineurin inhibitors or cyclophosphamide was not significantly associated with improved survival. However, the ever use of hydroxychloroquine (hazard ratio 0.59 (0.37-0.93); P=0.02) and azathioprine (hazard ratio 0.46 (0.28-0.75); P=0.002) was significantly associated with reduced mortality (41% and 54%, respectively) after adjustment for the propensity score and other confounding factors. A similar beneficial effect of hydroxychloroquine and azathioprine on survival was also observed in patients with lupus nephritis. Conclusions In this longitudinal cohort of Chinese SLE patients, the ever use of hydroxychloroquine and azathioprine was significantly associated with a probability of better survival. Treatment with high-dose prednisolone, cyclophosphamide, mycophenolate mofetil or the calcineurin inhibitors was not associated with long-term survival benefit.