Background: Patients with estrogen receptor (ER)- positive breast cancer and high-risk 21-gene recurrence score (RS) assay results benefit from chemotherapy, however some patients choose to decline chemotherapy. We evaluated factors associated with chemotherapy refusal by older women with high RS breast cancer and investigated the association of chemotherapy refusal with mortality. Methods: We used the National Cancer Database (2010-2017) to retrospectively identify women aged ≥65 years with ER-positive, HER2-negative, high RS (≥26) breast cancer. Women with Charlson Comorbidity Index ≥1, stage III or IV disease, or any unknown variables were excluded. Women with high RS who refused chemotherapy were compared to women with high RS who received chemotherapy. Refusal trends were analyzed using the Cochrane Armitage test. Factors associated with chemotherapy refusal were evaluated with a multivariable regression model. Overall survival (OS) by age and by treatment were evaluated with Kaplan-Meier and Cox proportional hazards modeling. Results: 6827 women met study criteria; 5449 (80%) received chemotherapy and 1378 (20%) refused. Relative to those who received chemotherapy, those who refused chemotherapy were older (median age 71 vs 69 years; p < 0.05), more often insured with Medicare (83% vs 80%; p = 0.05), were diagnosed more recently (2014-2017 vs 2010-2013, 67% vs 61%; p < 0.05), had lower grade tumors (grade 1 or 2 vs grade 3, 57% vs 48%; p < 0.05), more frequently had progesterone receptor-positive tumors (68% vs 63%; p < 0.05), and received radiation less frequently (67% vs 71%; p < 0.05). Chemotherapy refusal was significantly associated with increasing age (age 75-79 vs 65-74 OR 1.61, CI 1.4-1.85; age ≥80 vs 65-74 OR 3.24, CI 2.76-3.79) and more recent year of diagnosis (2014-2017 vs 2010-2013; OR 1.3 CI 1.14-1.48). Chemotherapy refusal was significantly associated with decreased 5-year OS for patients aged 65-74 years (92% vs 95%; p < 0.05) and patients aged 75-79 years (85% vs 92%; p < 0.05), but not for those aged ≥80 years (84% vs 91%; p = 0.07). Overall, when controlling for patient factors, hazard of death with chemotherapy refusal was significantly increased (HR 1.12, CI 1.04-1.2), but was not increased for women aged ≥80 years when stratified by age. Conclusions: Among healthy women aged ≥65 with high genomic risk ER-positive breast cancer, chemotherapy refusal increased with increasing age. Chemotherapy refusal was significantly associated with decreased OS in women aged 65-79, but did not impact OS in women aged ≥80. Lower use of chemotherapy in women ≥80 may demonstrate pragmatic decision-making between physicians and patients. Furthermore, the routine use of genomic assays may not be appropriate in this age group. More research is needed to determine why women aged 65-79 refuse chemotherapy, and whether patients remain satisfied with these choices.