Although the relationship of CYP2C19 polymorphism to citalopram disposition has been studied in healthy subject, this relationship in combination with dynamic effects (clinical adverse effect of citalopram) has not been well studied in patients. We carried out the present study to investigate the CYP2C19 genotype-phenotype relationship and potentially relate such relationship to the clinical effect (specifically adverse effects) of citalopram in Chinese patients who are known to have relatively high prevalence of poor metabolizers (PMs) of CYP2C19. Fifty-three Chinese adult patients were recruited. One to 2 blood samples at 4 to 24 hours postdose were collected after a minimum of 2 weeks of citalopram administration. The CYP2C19 genotypes were determined by polymerase chain reaction-restriction fragment length polymorphism, and the plasma concentrations of citalopram and desmethylcitalopram were determined by a liquid chromatography-tandem mass spectrometry method. The clinical adverse effects associated with citalopram were assessed according to Toronto Side Effects Scale (TSES). A population pharmacokinetic model was used to analyze the citalopram concentrations. Among 53 patients, 21 were homozygous extensive metabolizers (EMs) (CYP2C19*1/*1), 25 heterozygous EMs (CYP2C19*1/*2 or *1/*3), and 7 PMs (CYP2C19*2/*2 or *2/*3 or *3/*3). The metabolic ratios (plasma concentration of desmethylcitalopram to citalopram) were found to be 0.20 +/- 0.07, 0.15 +/- 0.05, and 0.07 +/- 0.03 in the homozygous EMs, heterozygous EMs, and PMs, respectively (P < 0.001, 1-way analysis of variance). On the basis of the results from our population pharmacokinetic modeling analysis, the citalopram oral clearances in the PMs were 42.9% and 33.3% (both P < 0.05) lower compared with the homozygous and heterozygous EMs, respectively. Statistically significant correlation was observed between the oral clearance and TSES scores in individual patients (rs = -0.37, P = 0.012). The mean TSES score also tended to be higher in PM than EM patients, but the difference was not statistically significant (P = 0.234). The study demonstrated a significant CYP2C19 genotype-phenotype relationship in Chinese patients receiving citalopram treatment. Such a relationship also tended to correlate with the clinical adverse effects of the drug. These results provide important pharmacogenetic implications for citalopram therapy in the Chinese population in whom relatively high frequency of CYP2C19 PM phenotype exists.