OBJECTIVE: Glucocorticoid (GC) tapering and withdrawal to reduce damage represents a key aspect of the European Alliance of Associations for Rheumatology (EULAR) SLE recommendations. However, optimal strategies for relapse-free GC cessation remain ill-defined. We characterised clinical predictors and their combined effect on flares in patients with SLE who discontinued GC. METHODS: Retrospective cohort of 324 patients with active SLE (PGA ≥1.5 and/or SLEDAI-2K ≥6) who received GC as part of treatment intensification (median follow-up 60 months). Survival and generalised linear models estimated SELENA-SLEDAI flare risks and their predictors. RESULTS: GCs were discontinued in 220 (67.9%) patients with 1-year risks for overall and severe flares of 50% and 25%, respectively (HR: 1.48
95% CI: 1.12 to 1.96 for overall flares
HR: 1.52
95% CI: 1.03 to 2.25 for severe flares, compared with non-withdrawers). Flare risk was lowered when GCs were ceased during remission (DORIS) or Lupus Low Disease Activity State (LLDAS
excluding remission) (HR for severe flares: 0.23
0.12 to 0.43 and 0.30
0.18 to 0.50, respectively), with each additional month in targets providing further protection. Hydroxychloroquine prevented total (HR: 0.37
0.26 to 0.53) and severe flares (HR: 0.33
0.21 to 0.52), while mycophenolate and azathioprine reduced overall flares. Prednisone tapering from 7.5 mg/day to 0 over >
6 months improved severe flare-free outcome (HR: 0.57
0.37 to 0.90). Random survival forests identified DORIS/LLDAS, hydroxychloroquine use and slow GC tapering as top predictors, whose coexistence reduced overall and severe flares by ~25 fold and ~50 fold, respectively. This combination reduced damage (IRR: 0.31
0.08 to 0.84) without inducing flares (IRR: 0.52
95% CI: 0.18 to 1.16) compared with GC non-withdrawers. CONCLUSION: Low or absent disease activity, slow tapering and hydroxychloroquine use minimise the risk of flares, facilitating GC discontinuation-a major goal in SLE.