Pediatric endocrinologists have long known that glucocorticoid replacement therapy, e.g. hydrocortisone given bid or even tid, does not reproduce the normal physiologic cortisol circadian pattern. Serum cortisol levels typically exceed physiologic levels shortly after dosing, long-term exposing patients to risks of overtreatment, such as growth retardation, weight gain, glucose intolerance, decreased bone density, and altered sleep pattern. At sleep onset, cortisol levels are low, while in the second half of the night cortisol levels are rising prior to waking. Increased cortisol levels at night are associated with light sleep and wakefulness. On the other hand, serum cortisol levels are typically subnormal before the next dosing, with attendant risks of hypoglycemia and other features of adrenal insufficiency. Studies in adults with adrenal insufficiency show impaired quality of life, with the most common complaint morning fatigue. In patients with congenital adrenal hyperplasia (CAH), glucocorticoid replacement is even more difficult, as normal or elevated ACTH levels will result in excess androgen production. Gauging replacement therapy is also difficult if more potent steroids (prednisone, dexamethasone) are used. This article by Debono et al. reviews evidence for normal cortisol production rate (5.7-7.4 mg/m2/day) and cortisol half-life. Debono et al. point out that the optimal time to administer hydrocortisome to reproduce the rising am levels is 0300 hrs, clearly not practical. The authors then discuss a “delayed, sustained release” hydrocortisone formulation (see also: Debono JCEM 94:1548-54, 2009). In the JCEM article, taking 15-20 mg of a “modified release” formulation at 2300 hrs and then 10 mg regular formulation at 0700 hrs did a reasonable job of reproducing the normal cortisol circadian pattern. Clearly, further trials in children are needed, but these new formulations show promise of improved treatment of adrenal insufficiency. Steve LaFranchi, MD