In this paper, the statistical relations between meteorological parameters and some pollutant ground level concentrations are presented. The daily average sulfur dioxide and smoke values were measured at five stations in Kayseri over 20 months. The model adopted for analysis differed from the power-law form selected by former investigators. Wind speed, degree-day temperature, the percentage of relative humidity, the previous day's pollution concentrations, and the amount of global solar radiation were the variables of multiple regression equations that were derived to calculate pollutant concentrations. The amount of cloud cover, however, had no important effect on the pollutant concentrations. The average variances of these regression equations were found as 84 and 75% for sulfur dioxide and smoke concentrations, respectively. The calculated pollutant concentrations utilizing forecast meteorological parameters reflected that the occurrence of high pollutant concentrations can be predicted.