1531 Background: Screening mammography reduces breast cancer specific mortality by allowing diagnosis at an earlier stage. Inconsistent guidelines and limitations in access to primary care are among factors that affect screening uptake. In this study, we reviewed patient characteristics, access to care, and uptake of screening mammogram (SMMG) prior to a breast cancer diagnosis at a safety-net system. Methods: Parkland Health (PH) is the safety-net system for Dallas County and is affiliated with the University of Texas Southwestern Simmons Comprehensive Cancer Center. Electronic medical records of patients with a new invasive breast cancer diagnosis between 2018 and 2019 at PH were reviewed and data on demographics and clinical presentation were collected. Patients were categorized based on SMMG uptake prior to their cancer diagnosis: never-screened, ever-screened (SMMG > 2 years prior to diagnosis) and recent-screened (SMMG within 2 years of diagnosis). Results: A total of 468 new breast cancer cases were identified (48.7% Hispanic; 32.9% Black). Of these, 12.2% were younger than 40 and 27.8% were 40-49. Payer mix included 64.7% uninsured and 34.8% Medicare/Medicaid. Half of the patients (50.6%) never had SMMG. Never-screened rate was higher among younger patients (98.2% in < 40; 53.8% in 40-49; and 39.4% in >50; p < 0.0001). Only 25.2% (118/468) of the patients had a SMMG within 2 years prior to their diagnosis. Among patients aged >40, 44.0% (181/411) were never-screened and 28.7% (118/411) were recent-screened. Screen-detection rate among patients >40 was 40.1% (165/411), of whom 22 (13.3% of screen-detects) were diagnosed with cancer on their first ever SMMG. Race/ethnicity, insurance status, and positive family history of cancer were not significantly different between never, ever, and recent-screened patients. Screening uptake correlated with nodal status, tumor size and stage at diagnosis. Patients who had a prior encounter at Parkland were more likely to have a SMMG. Patients established with Parkland primary care at diagnosis were more likely to have early-stage disease. In patients < 40, 24.6% presented with stage IV disease at diagnosis and 57.9% did not have a primary care provider. A positive family history of breast/ovarian cancer or any cancer was noted in 38.6% and 57.9% of patients < 40, respectively. Conclusions: In the cohort presented here, 27.8% of patients were between age 40-49, reflecting the impact of a gap in the guideline, which is now addressed in the most recent iteration of USPSTF guidelines. Connecting ED users to primary care, identifying higher risk individuals through family history, and increasing SMMG rate in the primary care setting are among strategies to reduce the rate of late-stage diagnosis of breast cancer, particularly among vulnerable patient populations.