It was aimed to characterize spatial variations of air pollutants in Marmara region, Turkey for determining contribution to air pollution status in this study. We used spatial data analysis for measured sulfur dioxide (SO2) and particulate matter (PM10) concentrations recorded in Marmara, which is the most industrialized region of Turkey. GIS technique was used for monitoring air pollution and spatial analyses of these pollutants measured with the period during between October 1, 2013 and March 31, 2014 known as winter (heating) season obtained from 61 air quality monitoring stations located in this region. Spatial distribution maps for these pollutants were generated to determine emission patterns for the study area with the aid of geostatistical techniques. Additionally standard and spatial regression models were employed on the measured emissions to reveal possible factors of air quality in the region using standard ordinary least squares (OLS) and spatially autoregressive (SAR) regression models. The two regression models revealed that all the four explanatory meteorological variables (i.e. temperature, wind speed, humidity and atmospheric pressure) used to depict the pollution levels in relation to air quality. After the definition of the final model parameters, the model was fit to the entire data set and the residuals were examined for the presence of spatial autocorrelation with Moran’s I. Compared to the OLS technique, SAR is found to be more appropriate when dependent variables exhibit spatial autocorrelation resulting in a valid model.