This study includes an approach to understand the factors affecting gas demand and to forecast gas consumption by multivariable regression analysis for the capital city of Ankara, Turkey. The process of the study is developing a statistical model and testing the model for the past years to understand how accurate it is. After obtaining the most reliable model, forecasting the gas consumption for the remaining days of 2002 and the year 2005 is performed. During the project, by the means of economical conditions, two scenarios, optimistic and pessimistic, are developed to get the idea of how the input variables are going to be changed in the following years. The model yields very satisfactory results and the range of gas consumption for the years 2002 and 2005 for the city Ankara is obtained.