BACKGROUND: Racial and ethnic disparities in receipt of recommended colorectal cancer screening exist; however, the impact of social determinants of health on such disparities has not been recently studied in a national cohort.OBJECTIVE: This study aimed to determine whether social determinants of health attenuate racial disparities in receipt of colorectal cancer screening.DESIGN: This was a cross-sectional telephone survey of self-reported race and ethnicity and up-to-date colorectal cancer screening. Associations between race/ethnicity and colorectal cancer screening were tested before and after adjustment for demographics, behavioral factors, and social determinants of health.SETTING: This was a nationally representative telephone survey of US residents in 2018.PATIENTS: The patients included were US residents aged 50 to 75 years.MAIN OUTCOME MEASURES: The primary outcome was up-to-date colorectal cancer screening status, according to 2008 US Preventive Services Task Force recommendations.RESULTS: This study included 226,106 respondents aged 50 to 75 years. Before adjustment, all minority racial and ethnic groups demonstrated a significantly lower odds of screening than those of non-Hispanic white respondents. After adjustment for demographics, behavioral factors, and social determinants of health, compared to non-Hispanic white respondents, odds of screening were found to be increased among non-Hispanic black respondents (OR, 1.10; p = 0.02); lower but attenuated among Hispanic respondents (OR, 0.73; p < 0.001), non-Hispanic American Indian/Alaskan Native respondents (OR, 0.85; p = 0.048), and non-Hispanic respondents of other races (OR, 0.82; p = 0.01); and lower but not attenuated among non-Hispanic Asian respondents (OR, 0.68; p < 0.001).LIMITATIONS: Recall bias, participant bias, and residual confounding.CONCLUSIONS: Adjustment for social determinants of health reduced racial and ethnic disparities in colorectal cancer screening among all minority racial and ethnic groups except non-Hispanic Asian individuals; however, other unmeasured confounders likely exist.