To the Editor: In follow-up to the report on the use of cyproheptadine in the treatment of Cushing's disease1 by Krieger and her co-workers and the subsequent correspondence2,3 we report a case that raises questions about the site and mode of action of cyproheptadine.A 61-year-old woman was seen in December, 1975, with hypokalemic coma and the clinical features of Cushing's syndrome. She had had increasing lassitude, weight gain, polyuria and ankle edema dating back to October, 1974. Plasma cortisol levels ranged between 40 and 50 μg per deciliter, with loss of diurnal variation. Urinary free cortisol levels were . . .