e18580 Background: Racial disparities in breast cancer have been extensively characterized in the literature. Access to screening mammography is a significant contributor to breast cancer disparities. Decreased access to mammography translates into disparities in breast cancer through delayed presentation to care. The USPSTF recommends biannual screening mammography of women 50-74 with average risk for breast cancer. Methods: We conducted a cross-sectional analysis of encounter-level data for Black and white female primary care patients aged 50-74 in the University of Washington Medicine system who were due for mammogram in 2019. Completion of steps of the mammography continuum (referral, scheduling, and completion of mammogram) were compared among Black and white women. Multivariable logistic regression was used to explore race and mammogram completion, adjusting for age, language, referral, insurance, clinical site, wellness visit utilization, and history of prior mammogram. Results: The study population comprised 18,156 women of whom 2,059 (11.3%) were Black and 16,097 (88.7%) were white. Among Black women, 26.8% were referred to screening mammogram, 12.9% self-referred, 39.3% were scheduled, and 21.4% completed their mammogram, compared to 21.1%, 20.6%, 41.4%, and 26.9% among white women respectively. The greatest attrition among Black women was in the step of completing a mammogram after it was scheduled, which was higher seen in white women at the same step. Adjusted analyses demonstrated an association between Black race and lower rates of screening mammography completion (OR 0.85, [95% CI 0.78-0.98], p = 0.02). Conclusions: Our analysis assessed racial disparities in steps of the screening mammography continuum in a large, diverse health system. Black race was associated with lower screening mammography completion after adjustment for several covariates. Provider-initiated referral was higher for Black women, while self-referral was higher for white women. Both Black and white women experienced highest attrition from no-show rates for scheduled mammograms, though attrition was greater for Black women. These findings have systems implications for future interventions, such as patient navigators or system-driven nudges, to mitigate disparities in breast cancer screening. [Table: see text]