Southeast Asia is rich in tropical forests and biodiversity but rapid deforestation and forest de- gradation have accelerated climate change and threatened sustainable development in the region. Carbon emission reductions through reducing deforestation and forest degradation, forest con- servation, sustainable management of forests, and enhancement of forest carbon stocks (REDD+) have been a focal topic of the climate change mitigation since the Bali in 2007. However, only a handful of studies exist so far on this important issue that are suitable to inform the debate with estimates of carbon stocks and emission reductions or removals as a result of REDD+. Our study attempts to analyze the potential emission reductions and removals for a 35-year period under the REDD+ scheme. We start by developing land use change and forest harvesting models that are used to estimate carbon stock changes in natural forests and forest plantations in Southeast Asia. Carbon emissions from deforestation and forest degradation of natural forests were 1865.1, 1611.4, and 1300.4 TgCO2 year−1, respectively. With a hypothetical carbon project of 35 years be- ginning from 2015, carbon emission reductions were estimated at 817.6 TgCO2 year−1, of which about 10% was from reducing forest degradation. Carbon removals due to increase of forest plan- tations were 76.3 TgCO2 year−1 but the removals could be much higher if there is a new definition on the eligibility of forest plantations. Summing up together, about 893.9 TgCO2 of carbon credits could be achieved from implementing carbon project in Southeast Asia or about US $6.6 billion annually between 2015 and 2050 if carbon price in 2012 is used. In addition to reducing emissions, there are other benefits from carbon project implementation. This study suggests that REDD+ has great potential for reducing carbon emissions and enhancing carbon stocks in the forests. Without financial incentives, carbon project would not happen and therefore climate change will continue to threaten future development.