Low-dose glucocorticoids (GC) achieve their action completely by classical genomic effects, mediated by the glucocorticoid receptor (GCR). In high doses of GC, nongenomic effects have also been found, but it is still unclear to what extent they contribute to a beneficial outcome. In this study, we present a determination of the number of lymphocyte GCR sites and the binding affinity in healthy children and children with autoimmune diseases. We further assess the effect of GC administration, especially of high-dose pulse therapy on the number of binding sites. The number of GCR sites per cell was analyzed with [(3)H]-dexamethasone radioligand binding assay and binding affinity (Kd given in nM) in peripheral blood mononuclear cells isolated from 48 healthy children and 35 patients. The patients were divided into three groups based on GC treatment: 0 mg/kg (group 1), 0.01-0.3 mg/kg orally (group 2), and 10-15 mg/kg i.v. pulse therapy (group 3) of prednisolone equivalent per day. Gender- and age-independent normal values of 4338 +/- 1687 sites/lymphocytes and Kd 6.7 +/- 2.2 nM were found. At 3463 +/- 1574, the number of receptor sites in patients without GC (group 1) was significantly lower than that of healthy volunteers (p < 0.05). In patients receiving GC treatment, this value was reduced to 2952 +/- 512 (group 2). Significant down-regulation to a minimum of 479 +/- 168 (group 3) was found after pulse therapy compared with untreated patients (p < 0.01). In pulse therapy, GC lead to a fast and dramatic receptor down-regulation. We suppose that the increase in therapeutic success of pulse-therapy may partly be mediated through additional nongenomic effects.