Background: Periostin has emerged as a novel biomarker in the pathogenesis of T helper 2-type allergic diseases in the last years. The aim of this study was to investigate the association of serum periostin levels with clinical features in children with asthma. Methods: Children with physician-diagnosed asthma who attended regularly to an outpatient pediatric allergy and asthma center were enrolled in the study along with control subjects. Asthma severity and control status of the patients were evaluated according to recent GINA guidelines. Results: A total of 158 children (125 with asthma and 33 age and sex-matched control subjects) with a median age of 10.2 years (range 5.9-17.0) were enrolled. Asthma severity was mild in 41 (32.8%), moderate in 63 (50.4%) and severe in 21 (16.8%) children. Children with asthma had significantly higher periostin levels than controls (53.1 ± 13.1 vs 43.0 ± 11.2 ng/mL; p < 0.001). The mean serum periostin levels of children with severe asthma (63.8 ± 10.8) were significantly higher than in children with moderate asthma (53.3 ± 12.7) and mild asthma (47.4 ± 11.1) (p < 0.001). Serum periostin levels were found to be significantly correlated with asthma severity (Spearman’s rho [r]=0.41, p < 0.001). Results of multivariable logistic regression analysis demonstrated an association between serum periostin levels and asthma severity in children (OR, 1.10; 95% CI, 1.04-1.15; p <0.001) Conclusion: Serum periostin levels may serve clinicians in identifying children with severe asthma.