Industrial goods are used daily by people worldwide.Although many changes that occurred as a result of industrial expansion were beneficial there are also a number of adverse effects, one being an increase in air pollution.The air quality index (AQI) provides us with a way of monitoring air pollution.In general 5 parameters are considered when calculating AQI.This study, only 3 parameters were chosen by statistical technique;O3, NO2 and SO2.These parameters were used to be principle variables in mathematical modeling, utilizing modified multiple regressions and finding unknown constants using by genetic algorithms technique compared with least square me method were considered.The data from the Pollution Control Department during 2003-2010 were used for predicting AQI in 2011.The results of this study show that the predicted value of the genetic algorithm technique is closer to the real value than the predicted value of least square method.