Within metropolitan areas, neighborhoods of growing immigrant settlement are becoming relatively less desirable to natives. We deploy a geographic diffusion model to instrument for the growth of immigrant density in a neighborhood. Our approach deals explicitly with potential unobservable shocks that may be correlated with proximity to immigrant enclaves. The evidence is consistent with a causal interpretation of an impact from growing immigrant density to native flight and relatively slower housing value appreciation. Further evidence indicates that these results are driven more by the demand for residential segregation based on ethnicity and education than by foreignness per se. (JEL I20, J11, J15, R23, Z13)