Abstract Background and Aims In lupus nephritis (LN), therapeutic schemas for remission induction are well known but data about duration of maintenance treatment are limited. There is no agreement between clinical practice guidelines about duration and withdrawal of therapy, and literature regarding this topic is scarce. Recently, a RCT has observed that immunosuppression discontinuation after 2‒3 years was related to LN relapse when compared with immunosuppression maintenance. However, immunosuppressive agents are related to severe adverse events and toxicity. We conducted this study with the aim to evaluate the incidence of LN relapse in a cohort of patients with discontinuation of immunosuppressive treatment after a first LN flare. We also aimed to determine factors associated to the presence of renal relapse in this population. Method Multicenter retrospective observational study including patients with biopsy proven LN who have received immunosuppressive treatment that was subsequently discontinued. Patients on prednisone at doses lower than 5 mg per day were allowed to be included. Patients were diagnosed between 1990 and 2018. The study was conducted in Spain. Relapse was defined according to Malvar et al criteria. Results 113 patients were included, mean age was 34.11 ±14.56 years-old and 85.8% were women. Serum creatinine at the time of LN diagnosis was 1.09±0.63 mg/dL, and proteinuria 3.16±2.53g/24h, ANA were positive in 85.84%, and anti-DNA Abs in 69.91%. Serum C3 and C4 levels were 72.51±44.06 and 13.04±11.27 respectively. 55.75% presented with class IV LN, 12.39% with class III or V LN, and 8.85% with class II. Mixed forms were less frequent. In a follow up period of 172.53±162.95 months, 17.7% (n = 20) presented a renal flare after immunosuppressive drugs withdrawal. Time from LN to relapse was 89.05±39.87 months, and time from immunosuppression discontinuation to relapse was 37.75±32.13 months. There were no differences in baseline characteristics between patients who relapsed and patients who did not, neither in histologic characteristics at the time of LN diagnostic biopsy except for glomerulosclerosis, that was less frequent in the relapse group (20% vs 34.41%, p = 0.0435). However, a tendency to a shorter time under immunosuppression in patients who relapsed was observed (73.55±79.06 vs 110.79±206.87 months). Data about treatment withdrawal in each group are summarized in Figure 1. Interestingly, there were no differences between both groups in steroids withdrawal (75% vs 78.49%,p = 0.7327) and hydroxychloroquine withdrawal (33.33% vs 21.74%, p = 0.3393). At the time of immunosuppression discontinuation, patients who relapsed presented lower serum C3 and C4 levels (78.79±24.89 vs 112.26±35.52, p = 0.0080 and 13.82±6.07 vs 23.09±12.56, p = 0.0333), but there were no differences in terms of renal function at this moment. At the end of follow up, patients who relapsed presented more albuminuria (227.87±259.79 vs 78.65±24.77, p = 0.0289) but creatinine levels were similar in both groups (0.84±0.30 vs 0.84±0.30, 0.5105). Only one patient died and two patients required renal replacement therapy, all in the no relapse group. Conclusion In a cohort of 113 patients with biopsy proven LN who were treated con immunosuppression and subsequently discontinued, 17.7% presented renal relapse at a mean time of 37 months after withdrawal. There was a tendency to a short time on immunosuppressive regimens in patients who relapsed. Also, lower serum C3 and C4 levels at the time of immunosuppression withdrawal were associated to renal relapse. At the end of a follow up of 14 years, patients who relapsed those who did not showed a similar maintained renal function but those who relapsed presented lower albuminuria.