Abstract:Abstract: Studying key factors of low carbon transformation from province level has great practical guiding significance for discovery carbon emission reduction potential. In this paper constructed dynamic panel data model and applied System-Generalized Method of Moments (SYS-GMM) to study influential factors affecting carbon emission intensity. Results are as follows: Firstly, lag variable of the explained variable (yt-1) is significant positive related to carbon emission intensity, dynamic locking-in effect of high carbon emission cannot be ignored. Secondly, GDP per capita and the industrial structure are significant positive factors, environmental Kuznets hypothesis is supported. Finally, R&D input has a significant negative related to carbon emission intensity,but urban population ratio, energy structure and foreign investment are not significant. Therefore, continuing optimize industrial structure and increasing investment in green technology research and development will be the main potential direction for China to accelerate regional low carbon transformation process。Simultaneously, how to relieve the dynamic locking-in effect of high carbon emission provinces is also a topic that needs attention for low carbon economy transformation in the future.