This article critically examines four specific aspects of Avner de Shalit’s book Cities and Immigration. First, it argues that the influx of cosmopolitan migrants, which de Shalit considers unproblematic for destination cities, may in fact pose a challenge to some cities’ ethos, and to the ethos of specific neighbourhoods within cities. Second, it contends that gentrification, contrary to what de Shalit suggests, may sometimes hinder rather than promote social mixing and migrants' integration. Third, it claims that most of the examples of mutual assimilation (de Shalit’s preferred model of inclusion) provided by de Shalit concern superficial interactions and exchanges between members of different groups. This raises both the empirical question of whether deeper forms of mutual assimilation are possible and the normative question of whether they are desirable. Fourth, it takes issue with two concrete policy recommendations advanced by de Shalit: the first involves assessing potential matchings between source and destination cities in order to advise migrants where to move; the second concerns prioritizing immigrants who might be better capable than others of facilitating the integration of future immigrants, due to their values, religion, or country of origin.