The relationship between the metabolism of the selective serotonin reuptake inhibitor citalopram and the sparteine and mephenytoin oxidation polymorphisms was studied in 24 healthy male volunteers, constituting panels of extensive metabolizers of sparteine and mephenytoin (n = 10), poor metabolizers of sparteine (n = 8), and poor metabolizers of mephenytoin (n = 6). Each subject was given 40 mg/day citalopram for 10 days and citalopram, and its des- and didesmethylmetabolites were assayed in serum and urine. Using a nonenantioselective analytical method (high-performance liquid chromatography), it was shown that the citalopram elimination partially depends on the mephenytoin oxygenase, since steady-state serum concentration, half-life, and area under the serum concentration/time curve for citalopram were significantly higher in poor metabolizers of mephenytoin than in extensive metabolizers of mephenytoin. Both citalopram total clearance and demethylation clearance (formation of desmethylcitalopram) were significantly lower in poor metabolizers of mephenytoin compared to extensive metabolizers (median 15.2 vs. 27.3 and 2.6 vs. 5.9 L/h, respectively). It was further indicated that the demethylation of desmethylcitalopram to didesmethylcitalopram depends on the sparteine oxygenase CYP2D6. Didesmethylcitalopram could virtually not be detected in any poor metabolizers of sparteine, contrasting measurable serum levels in all sparteine/mephenytoin extensive metabolizers. The demethylation clearance of desmethylcitalopram was significantly lower in sparteine poor metabolizers compared to extensive metabolizers (0.3 vs. 2.4 L/h, respectively). During administration of citalopram, there was a modest increase in sparteine metabolic ratio from median 0.31 to 0.80 in extensive metabolizers of sparteine, whereas the mephenytoin S/R ratio was unaltered during citalopram treatment. Both the sparteine and the mephenytoin oxidation polymorphism thus appear to contribute partially to the total pharmacokinetic variability of citalopram.