Climate is one of the environmental factors the change of which poses great threats to sustainable development. In recent decades, climatic changes that are mainly due to human activities have led to changes in different parts of the ecosystem. In this regard, general circulation models have been used by climate agencies to predict future climatic changes. In this study, the statistical downscaling model (SDSM) was used to predict precipitation and temperature at Babolsar and Gharakhil synoptic stations. The predictions were based on CanESM2 general circulation model scenarios. In the validation stage, the results of evaluating the accuracy of the SDSM model in rainfall simulation based on R2, RMSE, MAE, NSE at Babolsar and Gharakhil stations showed a great agreement between the simulated values and the observed values within the base period. Also, the values of precipitation and temperature predicted under RCP2.6, RCP4.5 and RCP8 scenarios at both stations in three periods including 2018-2039, 2038-2068 and 2069-2100 were compared to the values obtained within the baseline period (1987-2007). The results showed a decrease in the mean annual precipitation and a rise in the mean annual temperature under all the three RCP scenarios. For instance, in the period of 2068-2100, under the RCP 8.5 scenario (i.e. a pessimistic one) at Babolsar and Gharakhil stations, the average annual precipitation is decreased by 51.07% and 34.7% respectively. In contrast, the average annual temperature at those two stations undergoes the highest increase by 3.44°C and 3.86°C respectively, compared to the baseline period.