<p>Global climate change and subsequent increase in temperature and dry periods are expected to affect agricultural productivity and food security. Here, we modeled the impact of future climate on water yield and rice productivity across a range of elevations in two watersheds in Northern Iran using the AquaCrop model. After evaluating 19 Atmosphere-ocean general circulation models (AOGCMs) from the Intergovernmental Panel on Climate Change’s Fifth Assessment Report (AR5) using statistical metrics, the models with better performance at six stations were selected in predicting historical air temperature (T) and precipitation (P). Our results showed that, despite a general increase in minimum and max air temperature during near future period (2020–2040), spring months had the highest increase in Tmax, while the lowest increase occurred during the winter months for all the stations. The greatest change in precipitation was observed in summer months. Increases in the future T and P were predicted to be larger at higher altitudes. The annual trend of rice yield was increasing based on RCP4.5 scenario and a gradually increasing pattern in rice yield was revealed from the lowlands to uplands. Under the RCP8.5 scenario, the annual rice yield in most areas was predicted to decline in the future, except for the upland area. In the lowland areas, most of the AOGCMs predicted a decrease in the water productivity, but an increase in water productivity was observed in the upland areas. The information on water productivity would help devise strategic water management plans during periods of both water shortage and excess under a changing climate.</p>