We investigated the effect of dexamethasone (80 mg/kg per day for 2 days) and prednisolone (600 mg/kg per day for 2 days, equivalent to dexamethasone for glucocorticoid (GC) potency) on both pharmacokinetics and pharmacodynamics of midazolam (MDZ), a substrate for cytochrome P450 (CYP) 3A, in 8-week-old male Sprague-Dawley rats. Animals received a single injection of MDZ (pharmacokinetic study, 10 mg/kg; pharmacodynamic study, 55.5 mg/kg) in the tail vein 24 h after the last dose of GC or placebo. The elimination half-life (t(1/2)) and the area under the concentration-time curve of MDZ were significantly reduced by pretreatment with dexamethasone to 58.9% and 44.7% of the control value, respectively, and the clearance of MDZ was significantly increased by dexamethasone. Similar changes observed by prednisolone pretreatment did not reach significance. The t(1/2) of the dexamethasone pretreatment group (14.4+/-0.7 min) was significantly shorter than that of the prednisolone group (20.9+/-1.5 min). The amount of CYP3A2 protein and the activity of erythromycin N-demethylase were significantly increased by dexamethasone and prednisolone pretreatments, but dexamethasone showed a greater effect than prednisolone. Sleeping time was significantly shortened by dexamethasone and prednisolone pretreatment to 38.7% and 57.1% of control value, respectively. The current study demonstrates that the anesthetic effect of MDZ would be reduced in patients treated with dexamethasone or prednisolone, and that the CYP3A induction was greater by dexamethasone than by prednisolone, implying that the potency of CYP3A induction may differ among GCs even when GC activity is the same.