There is a lively debate about the causes of racial residential segregation. Counter to the racial-proxy hypothesis (Harris 1999, 2001), we argue that race, per se, continues to be influential when Whites make housing decisions. Using a survey-based experiment, we ask: Does information about neighborhood racial composition influence how Whites judge the quality of that neighborhood, quite apart from the actual characteristics of the homes located in it? A random sample of adults aged twenty-one and older in the Chicago and Detroit metropolitan areas watched videos embedded in a face-to-face interview. These videos portrayed neighborhoods ranging from lower working class to upper class. All respondents saw the same neighborhoods but were randomly assigned to see either (1) White residents, (2) Black residents, or (3) a mix of both White and Black residents. Respondents then evaluated the neighborhoods in terms of housing cost, property upkeep, safety, trajectory of housing values, and quality of the schools. Results show that Whites who saw White residents rated the neighborhood more positively on four of five dimensions than did Whites who saw the identical neighborhood with Black residents; racially mixed neighborhoods fell in between. In addition, Whites who endorsed negative stereotypes about Blacks were more likely to give low evaluations to neighborhoods with Black residents than were Whites who did not endorse stereotypes.