Data on flares in Asian patients with systemic lupus erythematosus (SLE) are scarce. Here, we aim to identify the baseline predictors of flares in a cohort of Southeast Asian patients with SLE. Consecutive adult patients with prevalent SLE according to the 1997 ACR or 2012 SLICC criteria were enrolled and followed three-monthly. Clinical and laboratory data were collected at every visit using a standardised protocol. Flares were defined using the SELENA-SLEDAI Flare Index (SFI). Baseline predictors of flare in patients with stable disease (SLE Disease Activity Index-2K (SLEDAI-2K) of ≤ 4) were determined using Cox proportional hazards. Of the 210 patients recruited, 148 (70.5%) were Chinese. The median (IQR) SLEDAI-2K at entry was 2 (0-4) and the median (IQR) disease duration was 10 (4.4-16.4) years. At baseline, 152 (72.4%) patients had stable disease. After a median (IQR) follow-up of 31.5 (24.1-36.3) months, 109 (51.9%) flared. Stable patients who flared tended to be in the lowest tertile of age (HR 3.08, 95% CI 1.72-5.48, p < 0.01), had thrombocytopenia (HR 5.01, 95% CI 1.32-18.99, p = 0.02), hypocomplementemia (HR 3.35, 95% CI 1.54-7.30, p < 0.01) and had the highest baseline prednisolone doses (HR 2.39, 95% CI 1.28-4.46, p = 0.01). Conversely, patients in the lowest tertile of disease duration tended not to flare (HR 0.41, 95% CI 0.21-0.80, p = 0.01). Flares are common in Asian SLE patients with initial stable disease. Close monitoring is needed for patients who are younger, with longer disease duration, thrombocytopenia, hypocomplementemia, or who required a higher baseline prednisolone dose.