The article focusses on the generating mechanisms of residential segregation for the demand side of housing markets, i.e., discriminatory residential preferences of inhabitants regarding the composition of their neighborhood. The data stem from an online survey among a random sample of the population of a mid-sized German city. In a vignette experiment, respondents were asked to rate example residential settings with respect to their attractiveness. The settings varied regarding the ethnic and religious composition of the neighborhood and other neighborhood characteristics that are positively or negatively related to residential attractiveness.We find that respondents have discriminatory residential preferences toward migrants and the presence of a Muslim community in the neighborhood. One-half of the migrant effect is mitigated if other positively connoted residential characteristics exist. We take this as an indication for statistical discrimination. This does not hold for the “Muslim community” effect. Discrimination gets stronger with higher levels of perceived economic group-threat from migrants. We further find evidence for a cultural group-threat and for the contact hypothesis: religious people are more discriminatory than nonreligious people, and real-life contact with migrants entails less discrimination.