Background: Screening for potentially inappropriate medication (PIM) use is recommended in elderly cancer patients receiving chemotherapy. However, few studies have examined the patterns and impact of PIM use in this population. Our objective was to determine predictors of PIM use and impact on outcomes in breast cancer patients receiving chemotherapy. Methods: We used data from the Surveillance, Epidemiology, and End Results database linked to Medicare claims. Our cohort included women 66 years and older with a diagnosis of Stage II/III breast cancer receiving adjuvant chemotherapy between 7/1/2007 and 12/31/2009. We used the National Center for Quality Assurance Drugs to Avoid in the Elderly list to define baseline PIM use as a dichotomous variable in the 4 months prior to diagnosis. Outcome measures included ER visits, hospitalization, or death within 6 months of diagnosis. We included age, race, stage, Charlson comorbidity, poverty level, and number of providers as covariates. We used Chi-square or Fisher’s exact test to determine associations of PIMs with covariates and outcomes, and multivariable logistic regression to determine the association of PIMs with outcomes. Results: 1595 patients met our criteria, of whom 339 (21.3%) had PIM use at baseline. In multivariate analysis, baseline PIM was associated with higher comorbidity (OR 1.33, 95% CI 1.0-1.76 for 1; OR 1.51, 1.07-2.13, for 2+), higher poverty level (OR 2.03, 1.43-2.88 for third quartile; OR 1.44, 1. -2.07, for highest quartile), and higher number of providers (OR 1.70, 1.18-2.44 for 4+ providers). 669 (41.9%) had an ER visit, unplanned hospitalization or died within 6 months after diagnosis. Baseline PIM was not significantly associated with the composite outcome in univariate analysis (OR 1.23, 95% CI, 0.97-1.57). Associations with outcome in the multivariate analysis included advanced stage (OR 1.48, 95% CI 1.19-1.82), higher comorbidity index (OR 1.67, 95% CI 1.23-1.26 for 2+ vs. 0) and baseline ER visits/hospitalizations (OR 1.82, 95% CI 1.33-2.49). Conclusions: We found no significant association between overall baseline PIM use and adverse outcomes. Ongoing analyses are evaluating the association between PIM use during chemotherapy and adverse outcomes.