In recent years, there has been a substantial amount of empirical work done on the causes of residential segregation. Nevertheless, better understanding of to what extent ethnic groups choose to live in the proximity of each other, or to what extent segregation is forced upon them is imperative. Prior research on self-segregation either focused on discovering underlying motivations for self-segregation, or the effect of stated preferences on observed patterns of segregation, whereas few studies directly link motivations, preferences, and segregation to one another in more detail. This article seeks to clarify mechanisms driving self-segregation, subsequently relating self-segregation to actual residential segregation. The results suggest that preferences for coethnic neighbors, driven mostly by interethnic prejudice, contribute to observed residential isolation to a certain extent. In some cases, perceived and experienced hostility and discrimination toward ethnic minorities stimulate self-segregation as well, while interethnic contact decreases it.Segregation has always been a key concept of urban sociology in studying the city and its inhabitants. To this day, questions about the causes of spatial segregation and its impact on social life are still dominant subjects in urban discourse. One of the major debates, which is still not battled out satisfactorily, deals with the question of to what extent segregation is voluntary or forced. In accordance with Chicago School reasoning, the spatial assimilation model posits that the level of residential segregation of immigrant groups reflects the level of assimilation into the host society in terms of economic success, language skills, and acculturation. However, this model fails to explain persistently high levels of segregation of mainly African Americans and therefore two alternative models have been developed. The first, known as the place stratification model, emphasizes the role of discrimination on the housing market, while the second, known as the preferences model, perceives self-segregation as a key factor in explaining patterns of segregation. Some argue that with declining institutional discrimination and whites' prejudice against blacks it ought to be blacks' preferences for black neighborhoods responsible for sustaining current patterns of segregation (e.g., Patterson, 1997;Thernstrom & Thernstrom, 1997). However, research on ethnic preferences is commonly unable to solve the choice-constraint issue, as few studies explicitly measured motivations and attitudes driving self-segregation. Because sometimes self-segregation can be directly linked to motivations, which favor the ethnic stratification model, its impact on actual residential isolation does not automatically imply that the preferences model holds true. Preferences for coethnic neighbors can, for instance, point toward the role of the neighborhood as a safe haven, in which ethnic minorities are able to "feel at home" due to the absence of discrimination practices that occur in white neighborhoods....