To identify the shortest duration of remission associated with improved outcomes in systemic lupus erythematosus (SLE). We studied 293 Caucasian patients with SLE during 7-year follow-up. Disease activity was assessed by SLE Disease Activity Index 2000 and damage by Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SDI). We defined three remission levels: complete, clinical off-corticosteroids, clinical on-corticosteroids (prednisone 1-5 mg/day). The effect of different durations of remission (1, 2, 3, 4 and ≥5 consecutive years) on damage was evaluated by multivariate logistic regression analysis. Among patients achieving 1-year (27 patients), 2-year (47 patients), 3-year (45 patients), 4-year (26 patients) remission, damage was similar irrespective of the level of remission achieved, whereas, among patients achieving ≥5-year remission (113 patients), damage was higher in those in clinical remission on-corticosteroids (p<0.001).In multivariate analysis, ≥2 consecutive year remission was protective against damage (OR (95% CI)): 2 years 0.228 (0.061 to 0.850); 3 years 0.116 (0.031 to 0.436); 4 years 0.118 (0.027 to 0.519) and ≥5 years 0.044 (0.012 to 0.159). Predictors of damage were cumulative prednisone dose ≥180 mg/month (3.136 (1.276 to 7.707)), antiphospholipid antibody syndrome (5.517 (2.092 to 14.546)), vasculitis (3.107 (1.030 to 9.307)) and number of flare/year (8.769 (1.692 to 45.449)). Two consecutive years is the shortest duration of remission associated with a decrease in damage progression in Caucasian patients with SLE.