Background: The landmark trials, TAILORx and RxPONDER have established Oncotype DX (ODX) as a prognostic and predictive tool in the management of estrogen receptor positive (ER+), early-stage breast cancer. However, recent studies have revealed existing racial disparities in the ODX test. These findings call into question whether ODX is generalizable to diverse populations. We evaluated the prognostic value of ODX RS at an urban, academic safety net hospital. Methods: We conducted a single-institution retrospective study of patients diagnosed with invasive breast cancer and had an ODX test performed between January 1, 2010 to December 31, 2021. Demographic, tumor and treatment characteristics were obtained. Chi-square and ANOVA were used to examine associated factors among recurrences and non-recurrences. Results: We identified 315 patients diagnosed with ER+ breast cancer with an ODX test obtained. Racial breakdown was 47% White, 35% Black and 18% Other. 69% of the patients had public or no insurance. 20% of the patients had a low RS (0-10), 62% had intermediate (11-25) and 18% were high (26-100). 264 patients had no evidence of recurrence, 15 patients had a local or distant recurrence and 36 were unknown. Recurrence was significantly associated with younger age (p = 0.04), higher histologic grade (P < 0.001, < 0.001), intermediate oncotype score (p = 0.002) and no receipt of chemotherapy (p = 0.014). Focused analysis of the recurrence cohort demonstrates a significant proportion with an Intermediate ODX RS classification (93.3%, P = 0.001). 53% had distant recurrence, 20% had received adjuvant chemotherapy and 93.3% were adherent to hormonal therapy. Conclusions: In a diverse patient population, recurrences were significantly more likely among younger patients with high grade tumors, intermediate RS and those who did not receive chemotherapy. Further investigation and understanding whether Oncotype DX is generalizable to diverse breast cancer patient populations is needed.