The effects of glucocorticoids (GC) are mediated by the activation of specific receptors that can be quantified in vitro by several laboratory tests. In vivo, other tests to determine GC sensitivity have been described, but only employing pharmacological doses. In this study, we used a very low dose of dexamethasone, an in vivo model to assess individual GC sensitivity. Fifty-five obese children and adolescents and 17 controls were studied. The patients were submitted to four 12-h urine collections, starting at 22:00 h; dexamethasone was administered orally at the end of the second urine sample. Patients were divided in the following groups: group Ob75 (n = 29) and the control group (n = 17) received dexamethasone 75 microg/m2, and group Ob150 (n = 26) received dexamethasone 150 microg/m2. Urinary cortisol was determined by RIA and expressed as microg/m2/12 h. All patients and controls showed a circadian rhythm before GC, which was maintained after dexamethasone only in controls. In the obese patients the circadian rhythm was abolished following both doses of dexamethasone, but more prominently with the dose of 150 microg/m2. In the obese group given 75 microg/m2, urinary cortisol inhibition was only observed in the first 12 h after dexamethasone, suggesting a partial and shorter suppression of the hypothalamic-pituitary axis. In both control and obese patients, the very low dose of dexamethasone was able to create a gradient of cortisol suppression that could be useful to identify an individual's sensitivity to glucocorticoids.