Istanbul has been affected by air pollution as a result of increasing population in the last two decades. Over the last 10 years, even with the thriving preventions of regional municipalities on the usage of natural gas instead of coal, the PM kind pollutions are not effectively controlled. To predict the dispersion of PM10 emissions ensuing from various air pollution sources-i.e. domestic, industries and vehicles, the Industrial Source Complex Short-Term Gaussian air quality dispersion model (ISCST3) was used. A grid configuration was set to cover a compact residential and industrial area of 4 x 4 km in the center of Istanbul. Domain area specific information, including meteorological data and emission rates of sources obtained by emission inventory, were put into the model. Comparison of predicted and measured 24-hr average PM10 concentrations was carried out to evaluate the model performance.