OBJECTIVE: To evaluate if fracture risk was increased in patients with Cushing's syndrome due to the increased endogenous cortisol production. DESIGN: Cohort. METHODS: A self-administered questionnaire was mailed to 125 patients with Cushing's syndrome diagnosed between 1985 and 1999 in Denmark. The response of each patient was compared with that of three age- and gender-matched control subjects randomly drawn among respondents to the same questionnaire from the background population. RESULTS: One hundred and four patients (83%) responded. The median age of the patients was 48 years (range 19-85 years). Sixty-eight had pituitary disease, 28 had adrenal disease, four had had both pituitary and adrenal surgery while four had not undergone surgery at the time of the study. The median time from diagnosis to surgery was 0.2 (range 0-3) years. Eighty-six percent were cured following surgery. There was an increased fracture risk within the last 2 years prior to diagnosis (incidence rate ratio 6.0, 95% confidence intervals (CI): 2.1-17.2). More than 2 years prior to diagnosis and following diagnosis there was no difference in fracture risk between patients and controls. The patients had more low-energy fractures than the controls (relative risk 5.4, 95% CI: 1.4-20.1). There was no difference in fracture risk between patients with adrenal or pituitary disease. CONCLUSIONS: Patients with Cushing's syndrome had an increased fracture risk in a narrow time interval before diagnosis, while no increase in fracture risk could be demonstrated after diagnosis and treatment.