Clinical and epidemiological studies, further supported by meta-analytic studies, indicate a possible association between chronicity (i.e., persistence or recurrence) of depression and hypothalamic pituitary adrenal (HPA) axis responsiveness to psychosocial stress. In the present study, we examined whether and how chronicity of depressive problems predicts cortisol responses to a standardized social stress test in adolescents. Data were collected in a high-risk focus sample (n = 351) of the Tracking Adolescents' Individual Lives Survey (TRAILS) cohort, a large prospective population study with bi- to triennial measurements. Depressive problems were assessed around age 11, 13.5, and 16. Cortisol levels were measured in saliva, sampled before, during, and after the Groningen Social Stress Test (GSST), to determine the cortisol response to psychosocial stress. The area under the curve with respect to the increase (AUCi) (i.e., change from baseline) of the cortisol response was used as a measure of HPA axis response. By means of linear regression analysis and repeated-measures analysis of variance, it was examined whether chronicity of depressive problems predicted the cortisol response to the GSST around the age of 16. Chronicity of depressive problems was significantly associated with cortisol stress responses. The relationship was curvilinear, with recent-onset depressive problems predicting an increased cortisol response, and more chronic depressive problems a blunted response. The results of this study suggest that depressive problems initially increase cortisol responses to stress, but that this pattern reverses when depressive problems persist over prolonged periods of time. (C) 2012 Elsevier Ltd. All rights reserved.