This study in human liver microsomes was undertaken to establish whether paroxetine stereoselectively inhibits the oxidative metabolism of metoprolol in vitro, and whether the in vivo observed magnitude of the paroxetine-metoprolol interaction was predictable from these in vitro data. Two distinct approaches were used: inhibitory effect of paroxetine on 1) the formation of alpha-hydroxymetoprolol and O-desmethylmetoprolol from the individual metoprolol enantiomers and 2) on the depletion of the enantiomers from the incubation mixture. Nonspecific binding of both metoprolol and paroxetine to human liver microsomes was also investigated. Whereas metoprolol displayed negligible binding, paroxetine was extensively bound to microsomal proteins. This was taken into account in order to obtain unbiased K(i) values and unbound concentrations of paroxetine. In the substrate depletion experiments, the intrinsic clearance (CL(int)) of (R)-metoprolol was larger than that of (S)-metoprolol. Paroxetine caused a concentration-dependent decrease in CL(int) of both enantiomers and abolished the stereoselectivity. In the metabolite formation experiments paroxetine did not stereoselectively affect alpha-hydroxylation, but preferentially inhibited the O-demethylation of the (R)-enantiomer versus the (S)-enantiomer. The use of unbound paroxetine concentrations in the two in vitro methods yielded comparable predicted increases in area under the curve (1.7-1.9 and 2.2-2.5 for (S)- and (R)-metoprolol, respectively) but underestimated the in vivo observed changes of about 7- and 10-fold, respectively. In conclusion, this study showed that paroxetine abolishes the stereoselective metabolism of metoprolol due to a stereoselective inhibition of the O-demethylation toward (R)-metoprolol. Furthermore, the extent of the in vivo metoprolol-paroxetine interaction was substantially underestimated by either one of the two in vitro approaches used when a competitive mechanism was assumed.