Introduction: Dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis is hypothesized to play a role in the pathogenesis of bipolar disorder (BD). Conflicting results have been reported when saliva or serum was used to measure cortisol levels. A recently developed method is to measure cortisol in scalp hair, with 1 cm of scalp hair representing 1 month. We studied whether there are differences in long-term hair cortisol levels between BD patients and healthy individuals and whether there are associations between hair cortisol and disease characteristics. Methods: Hair samples were collected in 100 BD patients and 195 healthy controls. Long-term cortisol levels were determined in 3 cm hair segments. Saliva samples were collected on two consecutive evenings. Documented disease characteristics were disease state, age of onset and psychiatric co-morbidity. Results: Hair cortisol levels were not statistically different in BD patients compared to healthy controls (p = 0.233) and were not associated with the disease state at the moment of sample collection (p = 0.978). In the subgroup of patients with age of onset >= 30 years, hair cortisol levels were significantly elevated compared to the subgroup with age of onset <30 years and to healthy controls (p = 0.004). Psychiatric co-morbidity was associated with elevated cortisol levels (44.87 versus 31.4 Conclusions: Elevated long-term cortisol levels might play a role in a subgroup of patients with BD. There may be differences in pathogenesis of younger and older onset BD suggesting two different disease entities. (C) 2012 Elsevier Ltd. All rights reserved.