Dextromethorphan is used widely in vivo to phenotype the polymorphically expressed cytochrome P450 (CYP) 2D6. Dextromethorphan is N-demethylated in vitro to 3-methoxymorphinan by human CYP3A4/5. We examined whether the dextromethorphan/3-methoxymorphinan urinary metabolic ratio (MR) could be used as an in vivo probe of CYP3A. Urinary excretion of 3-methoxymorphinan was excretion rate-limited in extensive metabolizers of CYP2D6, which necessitated a longer urine collection, 0 to 72 hours, to obtain true MR values for CYP3A. The urine excretion of dextromethorphan and 3-methoxymorphinan was delayed in poor metabolizers of CYP2D6 but appeared to be formation rate-limited. The delayed excretion in poor metabolizers necessitated longer urine collection intervals, 0 to 11 days, to estimate the true CYP3A MR and 0 to 8 days to estimate the true CYP2D6 MR. However, a 72-hour collection in poor metabolizers was used as an index of the true dextromethorphan/3-methoxymorphinan MR. Rifampin (300 mg b.i.d. for 7 days) significantly reduced the 0- to 72-hour dextromethorphan/3-methoxymorphinan MR consistent with an 830% (+/- 1808%) induction of CYP3A activity (n = 8), whereas erythromycin (250 mg q.i.d. for 7 days) significantly increased the dextromethorphan/3-methoxymorphinan MR, corresponding to a 34% +/- 44% inhibition of activity (n = 7) in extensive metabolizers and poor metabolizers. The changes in CYP3A activity were independent of CYP2D6 phenotype and were also observed after 24- and 48-hour urine collections in extensive metabolizers and poor metabolizers. In addition, MRs reflecting CYP2D6 and CYP3A were not significantly correlated. We conclude that the commonly used antitussive dextromethorphan can be used as an in vivo marker of CYP3A and CYP2D6 activity.