This paper examines the spatial relationship between the ambient air pollution level of an apartment and its property value in the housing market of South Korea. Using detailed transaction data for 2015–2018, we construct the air pollution index and estimate a two-stage spatial Durbin error model that controls for both direct and spillover effects. We find that, holding other factors equal, a 1% increase in the air pollution level can, on average, cause a decrease in the value of a local real property by 0.32% ($879). Spatially heterogeneous effects of air pollution on housing prices are investigated, and air pollution is found to have a more significant direct impact on the urban housing market than in rural areas. Moreover, rising air pollution levels in urban centers can raise housing prices in suburban and rural areas, suggesting a strong spillover effect of air pollution and potential migration towards better air quality. The findings in this paper have profound implications for analyzing the spatial impacts of air pollution on housing prices and urban development.