The genetically polymorphic cytochrome P450 2C9 (CYP2C9) metabolizes many important drugs. Among them, phenytoin has been used as a probe to determine CYP2C9 phenotype by measuring the urinary excretion of its major metabolite, S-enantiomer of 5-(4-hydroxyphenyl)-5-phenylhydantoin (p-HPPH). Phenytoin pharmacokinetic is also dependent on the activity of CYP2C19 and p-glycoprotein (ABCB1). To determine the influence of CYP2C9, CYP2C19 and ABCB1 genetic polymorphisms on phenytoin metabolism in a Black population, 109 healthy Beninese subjects received a single 300 mg oral dose of phenytoin. Blood was drawn 4 h after drug intake and urine was collected during the first 8 h. Plasma phenytoin and urine S- and R-enantiomers of p-HPPH were determined by high-performance liquid chromatography. Urinary excretion of (S)-p-HPPH [defined as urinary volume×(S)-p-HPPH urinary concentration] and PMR (defined as the ratio of p-HPPH in urine to 4 h phenytoin plasma concentration), both markers of CYP2C9 activity, were used to determine the functional relevance of new variants of CYP2C9 (*5, *6, *8, *9 and *11) in this population. Plasma phenytoin concentration was significantly associated with ABCB1 haplotype/genotype (P=0.05, Kruskal–Wallis test) and levels increased significantly in the genotype order: wild-type, T3421A and Block-2 genotypes (P=0.015, Jonckheere–Terpstra test). Urinary excretion of (S)-p-HPPH and PMR were significantly associated with the CYP2C9 genotype (P=0.001, analysis of variance (ANOVA) and P<0.0001, Kruskal–Wallis test, respectively) and decreased in the order: CYP2C9*1/*1, CYP2C9*1/*9, CYP2C9*9/*9, CYP2C9*1/*8, CYP2C9*8/*9, CYP2C9*9/*11, CYP2C9*1/*5, CYP2C9*6/*9, CYP2C9*1/*6, CYP2C9*8/*11, CYP2C9*5/*8 and CYP2C9*5/*6 (P<0.001, Jonckheere–Terpstra test). A combined analysis of CYP2C9, 2C19 and ABCB1 revealed that only ABCB1 predicted phenytoin concentration at 4 h and explained 8% of the variability (r2=0.08, P=0.04). On the other hand, only CYP2C9 was predictive for the urinary excretion of (S)-p-HPPH and PMR (r2=0.21, P=0.001 and r2=0.25, P<0.001, respectively). Furthermore, significant relation was found between urinary excretion of (R)-p-HPPH and CYP2C9 genotype (P=0.035) and levels significantly increased in the genotype order: CYP2C9*1/*9, CYP2C9*1/*1, CYP2C9*9/*11, CYP2C9*1/*8 and CYP2C9*1/*5 (P<0.001, Jonckheere–Terpstra test). In summary, the present study demonstrates that, in a Black population, CYP2C9*5, *6, *8 and *11 variants, but not CYP2C9*9, are associated with a decreased phenytoin metabolism. The data also confirm the limited contribution of MDR1 gene to inter-individual phenytoin pharmacokinetic variation.