The absorption of CO_2 by forest land is a critical mean of carbon capture and storage. Greenhouse gas emissions from deforestation and forest degradation have become the second major cause of global warming. Owing to these high forest emissions,in 2007 the United Nations Framework Convention on Climate Change (UNFCCC) introduced the Reducing Emissions from Deforestation and Degradation (REDD). An extension of this mechanism,titled REDD +,includes plans for forest protection,sustainable management of forests and enhancement of forest carbon sinks. The potential impact factors of reduction in forest carbon emissions of China are firstly analyzed, and key impact factors are identified by global regression model. Because the model is spatially unstable through statistical test, the geographical weighted regression model is used to analyze spatial heterogeneity of the key factors. The results show that reduction in forest carbon emissions of China is mainly affected by GDP per capita, natural population growth rate, population density, gross output value of agriculture and forestry. The GDP per capita with a negative effect on reduction in forest carbon emissions shows a decreasing trend from the northeast to the southwest,while the population density with a negative effect shows an increasing trend from the west to the east. The natural population growth has a positive effect on reduction in forest carbon emissions in the northeast region, and has a negative effect in the southwest region. Agricultural development can increase reduction in forest carbon emissions, and shows a decreasing trend from the east to the west. Forestry development plays an important promotional role in reduction in forest carbon emissions, showing a decreasing trend from the southwest to the northeast. In order to provide the decisionmaking basis for REDD + design of China in the future,some policy tools targeting on the spatial heterogeneity are finally proposed.