Black-White disparities in gifted enrollment persist across U.S. school systems. In this study, we examined whether these disparities depend on county-level rates of anti-Black bias. We drew data from the Civil Rights Data Collection, the Education Opportunity Project, and the Race Implicit Association Database. Based on a series of heteroskedastic fractional probit regression models, we found that county-level rates of anti-Black bias predict Black-White disparities in gifted and talented enrollment, with gaps being largest in counties with elevated rates of anti-Black bias and virtually non-existent in counties with low levels of anti-Black bias. These findings persist after accounting for achievement and other observable differences across counties, unobserved variation in state-level gifted and talented policies and are consistent for county-level estimates of both explicit and implicit measures of anti-Black bias.