Musk xylene (MX) (1,3,5-trinitro-2-t-butylxylene) is a nitromusk perfume ingredient that although uniformly negative in a battery of genotoxicity tests, produces a high incidence of liver tumors in mice. The purpose of this work was to characterize the profile and dose–response relationship of microsomal enzyme induction following exposure to MX. MX was dosed by gavage to male B6C3F1 mice for 7 days at 0, 1, 5, 10, 20, 50, 100, and 200 mg/kg after which microsomes were prepared. At 200 mg/kg, MX increased liver weight by about 65% and increased microsomal cytochrome P-450 content 2-fold over control. MX increased microsomal activity forO-dealkylation of 7-ethoxy and 7-methoxyresorufin 4- and 2-fold, respectively, and increased theN-demethylation of erythromycin approximately 2-fold. These results were generally consistent with increased CYP1A1, 1A2, and 3A protein levels determined by Western blotting. In contrast, whereas no increase inO-dealkylation of 7-pentoxyresorufin (PROD) was observed, MX treatment increased CYP2B protein levels about 25-fold over control at 200 mg/kg. Furthermore, a single dosage of MX (200 mg/kg) increased Cyp2b-10 mRNA to a maximal level and with a time course similar to phenobarbital (PB). To study inhibition of CYP2B enzymesin vivo,mice were treated with PB (0.05% in drinking water for 5 days), then given a single dosage of corn oil or MX (200 mg/kg) at 2 or 18 hr before necropsy. PB treatment increased PROD activity 25-fold, and at 2 hr after MX treatment (associated with peak plasma levels of MX), there was no change in the PB-induced PROD activity. However, at 18 hr, MX treatment decreased PROD activity by 90%. Despite thein vivoinhibition,in vitrostudies indicated that MX did not cause mechanism-based inactivation of CYP2B enzymes. The potential for nitroreduction of MX (catalyzed by anaerobic intestinal bacteria) to contribute to the inhibition of CYP2B enzyme activity was evaluated in a separate group of PB-induced mice that were dosed orally with a regimen of broad spectrum antibiotics (neomycin, tetracyline, and bacitracin) to reduce gut flora prior to administration of MX. In these animals, MX (200 mg/kg) did not inhibit PB-induced PROD activity. In summary, MX treatment produced general hepatic changes consistent with induction of CYP2B enzymes in mice and caused a large increase in CYP2B protein and mRNA levels. These data indicate that MX is a PB-like inducer of cytochrome P-450 enzymes and may cause liver tumors in a manner analogous to PB. However, no increase in CYP2B enzyme activity was observed, suggesting that MX or metabolites of MX also inhibit this enzyme. When the intestinal flora was eliminated by antibiotic treatment, MX no longer inhibited the CYP2B enzyme, indicating that anaerobic bacteria are capable of metabolizing MX, and suggesting that amine metabolites formed by nitroreduction are involved in the inhibition of mouse CYP2B enzymes.