Physicians' perspectives regarding the etiology of racial health differences may be associated with their use of race in clinical practice (race-based practice). This study evaluates whether attributing racial differences in health to genetics, culture, or social conditions is associated with race-based practice. This is a cross-sectional analysis, conducted in 2022, of the Council of Academic Family Medicine Education Research Alliance 2021 general membership survey. Only actively practicing U.S. physicians were included. The survey included demographic questions; the Racial Attributes in Clinical Evaluation (RACE) scale (higher scores imply greater race-based practice); and 3 questions regarding beliefs that racial differences in genetics, culture (e.g., health beliefs), or social conditions (e.g., education) explained racial differences in health. Three multivariable linear regressions were used to evaluate the relationship between RACE scores and beliefs regarding the etiology of racial differences in health. Of the 4,314 survey recipients, 949 (22%) responded, of whom 689 were actively practicing U.S. physicians. In multivariable regressions controlling for age, gender, race, ethnicity, and practice characteristics, a higher RACE score was associated with a greater belief that differences in genetics (β=3.57; 95% CI=3.19, 3.95) and culture (β=1.57; 95% CI=0.99, 2.16)-in but not social conditions-explained differences in health. Physicians who believed that genetic or cultural differences between racial groups explained racial differences in health outcomes were more likely to use race in clinical care. Further research is needed to determine how race is differentially applied in clinical care on the basis of the belief in its genetic or cultural significance.