6565 Background: Studies have documented disparities in breast cancer care for decades, but disparities in outcomes have not improved. Previous disparity investigations included relatively small, potentially biased samples of disadvantaged populations, or focused only on selected aspects of care. We sought to describe the relative importance of race/ethnicity versus Medicaid insurance status as determinants of suboptimal quality across the spectrum of care using two population-based samples of women with non-metastatic breast cancer. Methods: From two state registry datasets (NY and CA), we identified adult women diagnosed with stage 0-III breast cancer from 2004-2009. To these data, we merged enrollment and claims files from Medicaid and Medicare. Quality was assessed relative to 35 underuse and overuse measures derived from clinical practice guidelines. We compared measures across race/ethnic and Medicaid enrollment groups, and used logistic regression models to assess the relationships between race/ethnicity, Medicaid status, and quality relative to surgery, chemotherapy, radiation, and endocrine measures. Analyses were conducted in parallel for NY and CA, and for women <65 and ≥65 years old. Results: The sample, which comprised 80,079 from NY and 121,098 from CA, included 14%/6% blacks, 8%/15% Hispanics, 5%/12% API’s, and 19%/14% Medicaid enrollees, respectively from NY/CA. There was moderate-high correlation in measure performance across states and race/ethnic groups. Multivariable models demonstrated that blacks had lower odds of receiving recommended surgery versus whites, and whites had lower odds of receiving chemotherapy, whether recommended or not, versus other race/ethnic groups. Medicaid status was associated with lower odds of receiving recommended surgery, radiation therapy, and anti-estrogen therapy, but not chemotherapy, for patients <65 and ≥65. Conclusions: Medicaid status, a surrogate for socio-economic status, was associated with lower odds of receiving recommended care across a broad spectrum of breast cancer treatments. Understanding patterns of disparities will facilitate efforts to design and disseminate real world solutions that foster improvements in outcomes for the most disadvantaged populations.