Dataset accompanying the publication "The Impact of Objective and Subjective Measures of Air Quality and Noise on House Prices: A Multilevel Approach for Downtown Madrid" (Geographical Analysis 2013, 89-2, 127-148). Air quality and urban noise are major concerns in big cities. This paper aims at evaluating how they impact transaction prices in downtown Madrid. For that purpose, we incorporate both objective and subjective measures for air quality and noise and we use multilevel models since our sample is hierarchically organized into 3 levels: 5080 houses (level 1) in 759 census tracts (level 2) and 43 neighborhoods (level 3). Variables are available for each level, individual characteristics for the first level and various socio-economic data for the other levels. First, we combine a set of noise and air pollutants measured at a number of monitoring stations available for each census tract. Second, we apply kriging to match the monitoring station records to the census data. We also use subjective measures of air quality and noise based on a survey. Third, we estimate hedonic models in order to measure the marginal willingness to pay for air quality and reduced noise in downtown Madrid. We show that housing prices are better explained by subjective evaluation factors rather than objective measurements.