By establishing an evaluation system of low carbon economic growth efficiency (LCEGE), inclusive of carbon sink, a non-radical DEA model with slacks-based measure (SBM) was used to measure provincial LCEGEs in China during the period from 1998 to 2013. Based on this data, the spatial auto-correlation of Chinese LCEGE was analyzed. Finally, according to the 1998-2013 panel data from 30 provinces across the country, the paper built up a spatial panel data model to conduct empirical research on the factors influencing LCEGE. The research results show that during the sample period, China’s provincial LCEGEs differentiate from each other, and the average LCEGE in eastern coastal provinces phenomenally exceeds those of the inland provinces. In terms of the three regions, eastern China witnesses the highest LCEGE, which is followed by western China; and central China was last in the ranking. Moran’s I statistic result indicates that provincial LCEGEs have significant spatial auto-correlation and tend to cluster. Factors of industrial structure, energy consumption structure, and government policy exert a remarkably negative effect on LCEGE; while technological innovation, human capital, FDI, and foreign trade lend LCEGE a helping hand.