OBJECTIVES:A broad range of phenytoin doses is used in clinical practice, with the final 'maintenance' dose normally determined by trial and error. A common functional polymorphism in the SCN1A gene (one of the genes encoding the drug target) has been previously associated with maximum dose of phenytoin used clinically, and also maximum dose of carbamazepine, another antiepileptic drug with the same drug target. METHODS:We have related variation at the SCN1A IVS5-91 G>A polymorphism to maximum dose and to maintenance dose of phenytoin in 168 patients with epilepsy treated with phenytoin. We also related genotype to phenytoin serum levels at maximum dose and at maintenance dose of phenytoin. We genotyped the polymorphism using an Applied Biosystems Taqman assay. RESULTS:The polymorphism is associated with phenytoin serum concentration at maintenance dose (P=0.03). In a reduced cohort of 71 patients receiving phenytoin monotherapy this association is also significant (P=0.03). Neither association remains significant after Bonferroni correction for multiple testing. CONCLUSIONS:These results are not a replication of the original study. They do, however, support the hypothesis that this polymorphism influences the clinical use of phenytoin. They also demonstrate the utility of using multiple phenotypes in pharmacogenetics studies, particularly when attempting to separate pharmacokinetic and pharmacodynamic effects. As the SCN1A polymorphism affects phenytoin pharmacodynamics, it is particularly useful to obtain data on serum levels in addition to dose because association of a pharmacodynamic variant may be stronger with serum levels than dose as the serum level may eliminate or reduce pharmacokinetic variability.