BACKGROUND: Non-segmental vitiligo (NSV) is an autoimmune skin disorder that is difficult to determine disease activity/severity and thus to treat. Alarmins have emerged as promising biomarkers in various diseases, so further confirmation of their potential roles in NSV would be of considerable value. With the present work, we aimed to determine the serum levels of alarmins in patients with NSV, correlate these alarmins with disease activity and severity, and analyze the predictive value of the combination of these markers. METHODS: 104 NSV patients and 56 healthy controls were enrolled at the Xijing Hospital of Fourth Military Medical University between September 1, 2018, and June 30, 2019. The serum levels of alarmins (including IL-33, IL-1α, S100A9, S100A12, S100B, and HMGB1) were measured with enzyme-linked immunosorbent assays. The predictive performance of these biomarkers was evaluated with the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and other representative statistics. RESULTS: A total of 104 patients with NSV (mean [SD] age, 34.2 [13.0] years; 62 [59.6%] male) and 56 healthy controls (mean [SD] age, 34.8 [13.5] years; 34 [60.7%] male) were enrolled. For vitiligo diagnosis, S100B had the highest sensitivity (92.31%), whereas HMGB1 had the highest specificity (85.71%); the combination of IL-1α, S100B, S100A9, and HMGB1 increased the AUC value to 0.925, with a sensitivity of 87.50% and a specificity of 85.71%. Multivariate logistic regression analysis showed S100B (OR, 1.019; 95% CI, 1.002-1.038; P =0.03), S100A9 (OR, 1.002; 95% CI, 1.001-1.003; P<0.001), and HMGB1 (OR, 1.915; 95% CI, 1.186-3.091; P =0.008) were significantly associated with vitiligo activity. S100A9 had the highest accuracy in discriminating patients at the active stage from the stable stage, with an AUC value of 0.827. The combination of these alarmins had an AUC value of 0.860 to assess disease activity, with a sensitivity of 90.00% and a specificity of 72.97%. Furthermore, S100B (r=0.61, P <0.001), S100A9 (r=0.33, P <0.001), and HMGB1 (r = 0.51, P <0.001) levels were positively correlated with the affected body surface area (BSA) in NSV patients. CONCLUSIONS: Serum S100B, S100A9, and HMGB1 might be biomarkers for diagnosing and assessing the activity/severity of NSV, either used alone or in combination.