<p> Exemestane (EXE) is an endocrine therapy used to combat postmenopausal breast cancer. Several studies have reported substantial differences in clinical outcomes between EXE-treated patients, as well as inexplicable variability in serum concentrations of EXE and its major metabolite, 17&beta;-dihydroexemestane (17&beta;-DHE). For many pharmaceuticals, drug response is influenced by patient-specific genetic factors related to xenobiotic metabolism. Thus, it is possible that allelic variation in genes involved in EXE metabolism contributes to inter-individual differences in patient outcomes, possibly through differential EXE clearance or varied rates of metabolite formation. Historically, knowledge of phase I EXE metabolism has been extremely limited with significant ambiguity surrounding the identity of the specific hepatic enzymes involved. To address this gap in knowledge, <i>in vitro</i> studies were undertaken to better characterize hepatic phase I EXE metabolism and in particular, to assess the impact of genetic variation in drug-metabolizing enzymes on the production of EXE metabolites with inhibitory activity against aromatase. </p><p> The first part of this dissertation describes the identification of phase I EXE metabolites and details their capacity to suppress estrogen synthesis. Four metabolites, including 17&beta;-DHE, were detected in incubations of EXE with pooled human liver microsomes. 17&beta;-DHE and a novel metabolite, 17&alpha;-DHE, were formed in incubations of EXE with pooled human liver cytosol. The identities of phase I EXE metabolites were confirmed through comparison to reference compounds using UPLC/MS/MS. Anti-aromatase activity assays (AAA) revealed that 17&beta;-DHE is the only phase I EXE metabolite formed by human liver fractions that appreciably impedes estrogen formation. AAA also suggest that the inhibitory potency of EXE is unaffected by common nonsynonymous polymorphisms in aromatase. The latter half of this dissertation identifies hepatic enzymes that are likely to participate in phase I EXE metabolism. <i>In vitro </i> assays show that CBR1, AKR1Cs, and multiple hepatic CYP450s predominantly reduce EXE to 17&beta;-DHE with minor formation of additional inactive metabolites. Kinetic assays comparing 17&beta;-DHE formation by each wildtype enzyme to its common variant allozymes show that specific genotypes are associated with altered EXE metabolism <i>in vitro</i>. However, additional investigations are needed to determine the prognostic value of these associations for predicting <i> in vivo</i> EXE response.</p><p>