An alternative to intramuscular hydrocortisone self-injection was evaluated in healthy controls and in patients with adrenal insufficiency. Plasma cortisol concentrations were assayed after administration of 200 mg hydrocortisone by intramuscular injection (10 healthy subjects) or after insertion of an identical dose by rectal suppository (12 healthy subjects, 3 patients with adrenal failure). Plasma cortisol concentrations peaked at 1 hour (about 4000 nmol/l) following intramuscular injection and declined thereafter. After rectal administration, levels peaked at between 1 and 2 hours and persisted for 8 or more hours. Similar levels were achieved at 4 hours (about 1000 nmol/l) by both methods of administration. Self-treatment by rectal suppository may be useful in the prevention of Addisonian crises.