The dose-dependent first-pass metabolism of midazolam, a cytochrome P450 (CYP) 3A substrate, was separately estimated in the intestine and liver after administration into a jejunal loop of rats with differently modulated enzyme activity. Modulation of CYP3A enzyme activity of Sprague-Dawley rats was performed by pretreating the rats with inducers such as dexamethasone or by co-administering ketoconazole (an inhibitor) with midazolam. Bioavailabilities of midazolam administered into the jejunal loop at a dose of 10 micromol were 12% in untreated (control) rats, and 2% in dexamethasone-pretreated rats. Co-administered ketoconazole (2 micromol) significantly increased the bioavailability to 53% and 7%, respectively, in these rats. The intestinal first-pass metabolism of midazolam administered into the jejunal loop at a dose of 50 nmol in untreated and dexamethasone-pretreated rats, estimated by the mesenteric blood-collecting method in-situ, was 25% and 49% of absorbed amount, respectively. The intestinal first-pass metabolism of midazolam was reduced when ketoconazole (0.5 micromol) was co-administered or when the dose of midazolam was increased to 0.5 micrommol in these rats. Assuming that the contribution of intestinal first-pass metabolism could be negligible when midazolam was administered at a much higher dose of 10 micromol, the estimated hepatic first-pass metabolism of midazolam at a dose of 10 micromol in untreated rats, dexamethasone-pretreated rats, untreated rats given ketoconazole, and dexamethasone-pretreated rats given ketoconazole was, respectively, 86, 97, 46, and 92% of the amounts absorbed. In conclusion, the dose-dependent intestinal first-pass metabolism and the hepatic first-pass metabolism of midazolam in rats with differently modulated CYP3A activities was quantitatively estimated by in-vivo and in-situ absorption studies.