Background: We examined the longitudinal associations of autonomic nervous system (ANS) and hypothalamic-pituitary-adrenal (HPA) axis rest and reactivity measures with anxiety and depressive symptoms at one-year follow-up in children with anxiety disorders. Methods: In a clinical sample of 152 children with a primary DSM-IV anxiety disorder, aged 8 to 12 years, anxiety and depressive symptoms were assessed with the Multidimensional Anxiety Scale for Children and the Children's Depression Inventory at pre-treatment baseline and one year later, after treatment with cognitive behavioral therapy. At baseline, children participated in a 70 min stress task. Salivary cortisol was measured directly prior to and 20 min post stress task. Skin conductance level (SCL), heart rate and high frequency heart rate variability (HRV) were continuously measured during rest and the stress task. To investigate if rest or reactivity measures predicted anxiety and depressive symptoms at one year follow-up, linear regression analyses were conducted for rest and reactivity measures of SCL, heart rate, HRV and cortisol separately. Results: Higher SCL reactivity predicted less decrease of anxiety symptoms at one-year follow-up. Cortisol reactivity showed a weak association with depressive symptoms at one-year follow-up: lower cortisol reactivity predicted less decrease in depressive symptoms. Limitations: Only self-reported anxiety and depressive symptoms were used. However, all predictors were objective biological measures, hence there is no risk of shared method variance bias. Conclusions: These findings suggest that pre-treatment HPA and ANS responsiveness to stress are predictive biomarkers for a lack of symptom improvement in children with a clinical anxiety disorder. (C) 2016 Elsevier B.V. All rights reserved.