The main purpose was to assess the incidence and late outcome of Cushing's syndrome, particularly in Cushing's disease. Information for all patients diagnosed with Cushing's syndrome during an 11-yr period in Denmark was retrieved. The incidence was 1.2-1.7/million.yr (Cushing's disease), 0.6/million.yr (adrenal adenoma) and 0.2/million.yr (adrenal carcinoma). Other types of Cushing's syndrome were rare. In 139 patients with nonmalignant disease, 11.1% had died during follow-up (median, 8.1 yr; range, 3.1-14.0), yielding a standard mortality ratio (SMR) of 3.68 [95% confidence interval (CI), 2.34-5.33]. The SMR was partly attributable to an increased mortality within the first year after diagnosis. Eight patients died before treatment could be undertaken. The prognosis in patients with malignant disease was very poor. Patients in whom more than 5 yr had elapsed since initial surgery were studied separately, including a questionnaire on their perceived quality of health. In 45 patients with Cushing's disease who had been cured through transsphenoidal neurosurgery, only 1 had died (SMR, 0.31; CI, 0.01-1.72) compared with 6 of 20 patients with persistent hypercortisolism after initial neurosurgery (SMR, 5.06; CI, 1.86-11.0). In patients with adrenal adenoma, SMR was 3.95 (CI, 0.81-11.5). The perceived quality of health was significantly impaired only in patients with Cushing's disease and appeared independent of disease control or presence of hypopituitarism. It is concluded that 1) Cushing's syndrome is rare and is associated with increased mortality, in patients with no concurrent malignancy also; 2) the excess mortality was mainly observed during the first year of disease; and 3) the impaired quality of health in long-term survivors of Cushing's disease is not fully explained.