In this study, it was aimed to be predicted the maximum and minimum air temperature for the province Karaman by using the variables of 850hpa temperature, daily sunshine duration, mean temperature, relative humidity, and mean pressure. The prediction of maximum and minimum temperature belonging to the first months of four seasons was firstly made by 850hpa temperature and then, adding the variables of daily sunshine duration, mean temperature, relative humidity, and mean pressure to this in order, regression equations were obtained. In general, in the prediction of minimum and maximum air temperature, regression model, in which the independent variables of 850hpa temperature, daily sunshine duration, and mean temperature, gave better results and, the addition of independent variables of mean pressure and relative humidity to the model did not cause a significant increase in determination coefficients.