ACTH dependency of plasma androstenedione (A) and testosterone (T) was determined in normal and hirsute women by measuring the magnitude of change of A and T between the time of the cortisol (F) peak and F nadir in a diurnal study. There was a significant diurnal rhythm of A synchronous with F in both normal and hirsute women (P less than 0.01). Five of 12 hirsute women had a greater than normal diurnal swing of A (P less than 0.05), but only 2 of the 12 had a greater than normal diurnal swing of T. Responsiveness of A and T to 1/2 unit of intravenous ACTH was determined after dexamethasone 1 mg was given the night before. Plasma A and T were elevated in most of the hirsute women during acute ACTH suppression by dexamethasone, indicating ACTH-independent hypersecretion of androgens. Nine of 17 hirsute women had a greater than normal A response to ACTH (P less than 0.05). Those who had an exaggerated diurnal swing of A also had hyper-responsiveness of A secretion to ACTH. Only 2 hirsute women had an exaggerated T response to ACTH. Some T levels were decreased by ACTH. Seven of the 9 hiruste women who had an exaggerated A response to ACTH had a normal maximum F response, but a greater than normal 17-hydroxy-progesterone (17-OHP) response to ACTH with a high 17-OHP to F ratio, suggesting they have a mild but compensated reduction in 21-hydroxylase or 11beta-hydroxylase activity. Two women with hyper-responsiveness of A secretion had low F and 17-OHP responses to ACTH suggesting reduced C21 but intact C19 3beta-hydroxysteroid dehydrogenase-delta-5,-4 isomerase activity. These apparent reduced enzyme activity may not be congenital, but induced by an altered hormonal milieu such as an abnormal androgen-estrogen ratio. It is concluded that ACTH uniformly stimulated A secretion but not T secretion and that approximately 50% of the hirsute women had ACTH-dependent hypersecretion of A, but most of these also had concurrent ACTH-independent hypersecretion of androgens.