OBJECTIVE: To assess the relative contribution of intestinal and hepatic CYP3A inhibition to the interaction between the prototypic CYP3A substrates midazolam and clarithromycin. METHODS: On day 1, 16 volunteers (eight men and eight women; age range, 20 to 40 years; weight range, 45 to 100 kg) received simultaneous doses of midazolam intravenously (0.05 mg/kg over 30 minutes) and orally (4 mg of a stable isotope, 15N3-midazolam). Starting on day 2, 500 mg clarithromycin was administered orally twice daily for 7 days. On day 8, intravenous and oral doses of midazolam were administered 2 hours after the final clarithromycin dose. Blood and urine samples were assayed for midazolam, 15N3-midazolam, and metabolites by gas chromatography-mass spectrometry. RESULTS: There was no significant (p > 0.05) difference in the urinary excretion of 1'-hydroxymidazolam after intravenous and oral dosing on day 1 or day 8, indicating that the oral dose was completely absorbed into the gut wall. The oral clearance of midazolam was found to be significantly greater in female subjects (1.9 +/- 1.0 versus 1.0 +/- 0.3 L/hr/kg; p < 0.05) than in male subjects but not systemic clearance (0.35 +/- 0.1 versus 0.44 +/- 0.1 L/hr/kg). For women not receiving oral contraceptives (n = 6) a significant gender-related difference was observed for systemic and oral clearance and for area under the curve and elimination half-life after oral administration. A significant (p < 0.05) reduction in the systemic clearance of midazolam from 28 +/- 9 L/hr to 10 +/- 3 L/hr occurred after clarithromycin administration. Oral midazolam availability was significantly increased from 0.31 +/- 0.1 to 0.75 +/- 0.2 after clarithromycin dosing. Likewise, intestinal and oral availability were significantly increased from 0.42 +/- 0.2 to 0.83 +/- 0.2 and from 0.74 +/- 0.1 to 0.90 +/- 0.04, respectively. A significant correlation was observed between intestinal and oral availability (n = 32, r = 0.98, p < 0.05). After clarithromycin administration, a significant correlation was observed between the initial hepatic or intestinal availability and the relative increase in hepatic or intestinal availability, respectively. Female subjects exhibited a greater extent of interaction after oral and intravenous dosing than male subjects (p < 0.05). CONCLUSION: These data indicate that in addition to the liver, the intestine is a major site of the interaction between oral midazolam and clarithromycin. Interindividual variability in first-pass extraction of high-affinity CYP3A substrates such as midazolam is primarily a function of intestinal enzyme activity.