Sale prices of residential properties are not necessarily affected by objectively measured property attributes, but rather by attributes which sellers and buyers perceive as factual. To verify this assumption, we examined the effect of air pollution on market prices of 926 housing units sold between 1998 and 2007 in the Greater Haifa Metropolitan Area, Israel. First, we compared the ambient levels of SO2 and PM10 pollution, measured by air quality monitoring stations, with subjective evaluations of air pollution levels in the city neighborhoods, obtained via interviews of local residents. These two types of air pollution estimates were then used in the multivariate analysis as property price predictors. As found, subjective evaluations of local air quality performed significantly better in explaining the variation of local apartment prices than objectively measured air pollution levels.