The management of congenital adrenal hyperplasia due to 21-hydroxylase (CYP21) deficiency requires glucocorticoid substitution with oral hydrocortisone given twice or thrice daily. In paediatric practice little is known of the bioavailability of oral hydrocortisone tablets used in these patients. The aim of this study was to assess the bioavailability of oral hydrocortisone and to evaluate current replacement therapy in the light of cortisol pharmacokinetic properties. We determined the bioavailability of hydrocortisone following oral and intravenous administration in sixteen (median age: 10.9 years, range: 6.0-18.4 years) adequately controlled CYP21 deficient patients. Serum total cortisol concentrations were measured at 20-min intervals for 24 h while patients were on oral substitution therapy, and at 10-min intervals for 6 h following an intravenous bolus of hydrocortisone in a dose of 15 mg/m(2) body surface area. The area under the serum total cortisol concentration versus time curve (AUC) following oral and intravenous administration of hydrocortisone was calculated using the trapezoid method. The bioavailability was estimated by dividing the corrected for dose AUC after oral hydrocortisone administration by the corrected for dose AUC after the intravenous hydrocortisone administration and was exemplified as a percentage. After oral administration of hydrocortisone in the morning, median serum total cortisol concentrations reached a peak of 729.5 nmol/l (range: 492-2520 nmol/l) at 1.2 h (range: 0.3-3.3 h) and declined monoexponentially thereafter to reach undetectable concentrations 7 h (range: 5-12 h) after administration. Following administration of the evening hydrocortisone dose, median peak cortisol concentration of 499 nmol/l (range: 333-736 nmol/l) was attained also at 1.2 h (range: 0.3-3.0 h) and subsequently declined gradually, reaching undetectable concentrations at 9 h (5-12 h) after administration of the oral dose. After the intravenous hydrocortisone bolus a median peak serum total cortisol concentration of 1930 nmol/l (range: 1124-2700 nmol/l) was observed at 10 min (range: 10-20 min). Serum cortisol concentrations fell rapidly and reached undetectable levels 6 h after the hydrocortisone bolus. The absolute bioavailability of oral hydrocortisone in the morning was 94.2% (90% confidence interval (CI): 82.8-105.5%) whereas the apparent bioavailability in the evening was estimated to be 128.0% (90% CI: 119.0-138.0%). We conclude that the bioavailability of oral hydrocortisone is high and may result in supraphysiological cortisol concentrations within 1-2 h after administration of high doses. The even higher bioavailability in the evening, estimated using as reference the data derived from the intravenous administration of hydrocortisone bolus in the morning, is likely to reflect a decrease in the hydrocortisone clearance in the evening. Decisions on the schedule and frequency of administration in patients with congenital adrenal hyperplasia should be based on the knowledge of the bioavailability and other pharmacokinetic parameters of the hydrocortisone formulations currently available.