The N-acetylation of the noncompetitive AMPA antagonist talampanel (TLP) represents a route of varying significance in various species. For a detailed analysis in humans, plasma concentrations of TLP and its N-acetyl metabolite (NAc–TLP) were measured for up to 48 h after administration of a single oral dose of 75 mg in 28 healthy volunteers following genotyping for the N-acetyltansferase NAT2 isozymes (alleles NAT2*4, *5, *6, and *7). Unified parent–metabolite pharmacokinetic (PK) models that allowed three different rates of acetylation were used to simultaneously fit plasma levels for both the parent drug and its metabolite following genotype-based classification as slow, intermediate, or fast acetylator. A perfect correspondence was found between the phenotype inferred from genotyping and the phenotype determined by using plasma metabolite-to-parent molar ratios indicating that this route of metabolism is indeed mediated by NAT2. Linear parent-metabolite PK models (first-order input, first-order elimination through two parallel routes one of which is through a metabolite with polymorphic rate of formation) gave adequate and sufficiently consistent fit. Parameters obtained suggest that for TLP in humans, N-acetylation represents only about 1/4th of the total elimination even in true (*4/*4 homozygous) fast acetylators, acetylation is about 8–12 times faster in fast and 3–6 times faster in intermediate acetylators than in slow acetylators, and the N-acetyl metabolite is eliminated faster than the parent drug. Such PK models can provide quantitative estimates of relative in vivo metabolism rates for routes catalyzed by functionally polymorphic enzymes.