To evaluate whether the occurrence or severity of gingival hyperplasia is associated with liver function test results or phenytoin metabolism. Prospective analysis. University-affiliated medical center in Taipei, Taiwan. Sixty-six patients (mean age 37.9 yrs) with epilepsy who were receiving phenytoin for more than 1 year. Intervention. Four blood samples were drawn from each patient for liver function testing, concentrations of phenytoin and its metabolites R-5-(4'-hydroxyphenyl)-5-phenylhydantoin (R-HPPH) and S-HPPH, and genotyping of cytochrome P450 (CYP) 2C9 and 2C19. Plasma concentrations of phenytoin and its metabolites were determined by a high-performance liquid chromatography method. The CYP2C9 and CYP2C19 genotypes were analyzed by polymerase chain reaction-restriction fragment length polymorphism analysis. Conventional liver function assays and a quantitative liver function test--galactose single-point (GSP) measurement--were performed. Statistical analyses were performed to evaluate the association between liver function test results as well as metabolic phenotype and the occurrence and severity of gingival hyperplasia. Among liver function tests, only GSP levels showed a significant difference between patients with and those without gingival hyperplasia. Patients with an elevated GSP level (> or = 280 microg/ml) had a significantly higher odds ratio (OR 4.51) for the occurrence of gingival hyperplasia. In addition, increased R-HPPH (OR 1.02) and phenytoin (OR 1.09) concentrations were associated with an increased occurrence of gingival hyperplasia. However, only increased GSP and R-HPPH concentrations had significantly higher ORs (2.84 and 1.02, respectively) associated with the severity of gingival hyperplasia. Although mean +/- SD plasma R-HPPH concentration was significantly lower in CYP2C19 poor metabolizers compared with CYP2C9 and CYP2C19 extensive metabolizers and CYP2C9 poor metabolizers (30.38 +/- 16.73 vs 68.22 +/- 44.75 and 78.95 +/- 51.67 microg/ml, respectively), no significant association between genotype and gingival hyperplasia was found. Increased GSP, phenytoin, and R-HPPH concentrations were associated with increased occurrence of phenytoin-induced gingival hyperplasia; only increased GSP and R-HPPH concentrations were associated with increased severity of this adverse effect.