Extreme climate events have increased in terms of their amplitudes, frequency and severity, greatly affecting ecosystem functions and the balance of the global carbon cycle. However, there are still uncertainties about how extreme climate change will affect tree growth. This study characterized the responses of tree growth to extreme climate on the northeastern Tibetan Plateau from 2000 to 2020. Meanwhile, a back propagation neural network was used to predict tree growth trends under two future emission scenarios from 2020 to 2050. This study revealed that: (1) the tree-ring width index (RWI) showed a decreasing trend (- 0.04/decade) in the eastern region, but the enhanced vegetation index (EVI) showed an increasing trend (0.05/decade) from 2000 to 2020. While both RWI and EVI in the middle and western regions showed increasing trends. (2) The responses of RWI and EVI to extreme climate were regionally asymmetric. In the eastern region, extreme precipitation inhibited tree radial growth, while extreme warm nights promoted tree canopy growth. In two other regions, both extreme precipitation and extreme warm nights promoted tree growth. (3) The model predicts that there was no significant change in RWI and EVI in the western region, but both RWI and EVI showed an increasing trend in the middle and eastern regions under the low emission scenario. Under the high emission scenario, the growth of tree stem and canopy in all three regions shows a general decreasing trend. The results of this study both improved the understanding of the differences in carbon allocation between tree stem (RWI) and canopy (EVI) and identified vulnerability thresholds for tree populations.