6558 Background: We studied care for early-stage breast cancer in Appalachia, a region with health infrastructure, socioeconomic (SES) and geographic disparities. Methods: Cases of stage I-III breast cancer diagnosed 2006-2008 were identified from cancer registries of KY, NC, OH, and PA and linked to Medicare data. Guideline concordance was studied in eligible groups, as follows: endocrine therapy for hormone receptor positive cancer (n=1429); and radiation (RT) use after breast conserving surgery (BCS) divided into two groups - age 70 years and older with ER/PR+, <2 cm, node negative tumors where it may have been acceptable to forgo RT (OptRT, n=1108) and all other cases (IndRT, n=1422). Multivariate (MV) and univariate analyses were performed. Covariates included age, state, Appalachian Regional Commission (ARC) economic status, Commission of Cancer (CoC) status, state, access to care, number of beds, surgery facility ownership, volume, and chemotherapy/radiation offered, provider graduation year and volume, Charlson comorbidity, diagnosis year, Medicaid/Medicare dual status, histology, tumor size, tumor sequence, positive lymph nodes, ER/PR status, stage, herceptin use, and BCS/mastectomy indicator. Results: Mean age was 74 years and 97% were white. Guideline-concordance was 76% for endocrine therapy, 83% for IndRT, and 77% for OptRT. Younger age predicted higher concordance in all groups. Endocrine therapy use was lower in NC vs PA (OR 0.60; 95% CI 0.41-0.88) and greater for cases whose provider graduated in years 1984-1988, vs. 1989+ (1.55; 1.06-2.29). In IndRT, provider volume in the 3rdquartile vs. highest quartile predicted increased radiation use (2.36; 1.46-3.81). In OptRT, less receipt of radiation was predicted by residence in NC vs. PA (0.26; 0.18-0.48), and competitive ARC class vs. transitional (0.60; 0.36-0.99). Conclusions: Within Appalachia, there are SES and provider characteristics that are associated with use of guideline concordant care.