In this research a multiple linear regression model is developed to predict natural gas consumption in Sarajevo Canton using meteorological parameters and air pollutants as variables. Data on natural gas consumption were collected from the only local natural gas distributor company Sarajevo gas, and meteorological data were collected from the Federal Hydro meteorological Institute of Bosnia and Herzegovina. The findings of this research show that temperature, atmospheric pressure, and relative humidity significantly influence gas consumption, explaining 91,50% of its variability. Higher values of these meteorological parameters correspond to lower gas consumption. The most important variable in the model is temperature. This research provides important insights for managing natural gas consumption in Sarajevo Canton during complete heating season as a whole in response to changing weather conditions.