Objective: Hair cortisol (HF) and cortisone (HE) measurements reflect tissular exposure to cortisol over months and are increased in overt Cushing’s syndrome (CS). No data is available in mild CS. We compared the diagnostic performance of HF and HE between patients with overt or mild CS. Design: Single centre retrospective study. Methods: HF&HE were measured by LC-MS/MS in 48 consecutive adult females with Cushing’s disease (CD), ectopic ACTH syndrome, secreting adenomas and carcinomas, and adrenal incidentalomas. All had impaired dexamethasone suppression tests. Overt CS (n = 25) was diagnosed in front of specific symptoms, a mean UFC (>1.5 ULN) and increased midnight serum cortisol or salivary cortisol. Mild CS (n = 23) was diagnosed in patients lacking specific symptoms and displaying at least one additional biological abnormality including mildly increased UFC (≤1.5 ULN), increased midnight serum cortisol or salivary cortisol and suppressed plasma ACTH in patients with adrenal tumours. In this study, 84 healthy subjects and obese patients served as controls. Results: HF and HE showed roughly similar performance in overt CS (92 and 100% sensitivity, 91 and 99% specificity, respectively). HF and HE were lower in mild CS but higher than in controls (P < 0.01). HE was correlated with midnight serum cortisol (P < 0.02) and volume of adrenal incidentalomas (P < 0.04) but not with UFC. HF and HE had 59% and 68% sensitivity, and 79 and 94% specificity, respectively, for the diagnosis of mild CS. Contrary to UFC, both HF and HE were in the range of overt CS in 11/23 patients with mild CS. Patients with mild CS and increased HE required more antihypertensive treatments and showed worser lipid profiles than patients with normal HE. Conclusions: HF and HE measurement performed better in overt than in mild CS but is a useful adjunct to diagnose mild CS and to identify adrenocortical incidentalomas responsible for excessive cortisol exposure.< Réduire