The purpose of this study is to investigate 1-year adherence rates to aromatase inhibitors (AI) and to determine risk factors for non-adherence among commercially insured post-menopausal breast cancer patients. A retrospective cohort of 13,593 commercially insured breast cancer patients with a prescription claim for an AI therapy (exemestane, anastrozole, and letrozole) in 2006 were identified using the MarketScan Commercial Claims and Encounters Database. Adherence was calculated by the medication possession ratio (MPR) for a 1-year period following the initial claim in 2006. The main outcome variable was non-adherence (<80% MPR) to AI therapy. Multivariate logistic regression was used to determine predictors of non-adherence. Over a 1-year period, 23% of patients were non-adherent with their AI therapy. AI non-adherence was associated with younger age, out-of-pocket costs ≥$30 per AI prescription, as well as with measures during the 12-month period prior to the initial 2006 AI claim of lower patient out-of-pocket total pharmacy costs, no mastectomy, and higher Charlson Comorbidity Index. Non-adherence to AI is a common occurrence. A number of factors were identified that influence patience non-adherence. These factors can be used to identifying patients at increased risk for non-adherence to better target and intervene to support medication taking behaviors.