<h3>Introduction</h3> Airflow limitation (AFL) has been shown to predict poor asthma outcomes across maturation. The objective of our study is to identify significant predictors of AFL in children with severe asthma from school-age to adolescence. <h3>Methods</h3> Children ages 5-18 years with treatment-refractory asthma underwent spirometry and diagnostic bronchoscopy. AFL was defined as pre-bronchodilator FEV<sub>1</sub>/FVC <0.85 or <90% predicted. Demographics, treatments, and blood and lung markers of inflammation were analyzed with between group non-parametric tests and logistic regression models. <h3>Results</h3> 226 children were included in this analysis. Nearly one-half (n=105, 46%) had AFL. There was no difference in age, gender, BMI, or age of symptom onset between groups. Compared to children without AFL, those with AFL were more likely to have previous ICU admission for asthma, treatment with higher-dose inhaled corticosteroids (ICS), and reported black race. Children with AFL had significantly higher total blood and bronchoalveolar lavage (BAL) eosinophil count, total IgE, and greater number of positive allergen specific serum IgE. In younger children, those with sensitization to ≥4 allergens had greater airflow limitation but the opposite relationship was seen in adolescent children. Similarly, the effects of higher total IgE (>130 IU/mL) on airflow limitation were most prominent in younger children. <h3>Conclusion</h3> Significant predictors of airflow limitation in children with severe asthma include black race, previous ICU admission for asthma, higher-dose ICS treatment, and increased absolute blood and BAL eosinophil count. AFL is driven in part by allergic sensitization in the younger children whereas in older children this associated more with the absence of allergic sensitization.