A benchmark map of forest carbon stocks in tropical regions across three continents has been developed to provide a reliable estimate of biomass carbon stocks for REDD (Reducing Emissions from Deforestation and Forest Degradation) assessments. The map covers 2.5 billion hectares of forests in Latin America, sub-Saharan Africa, and Southeast Asia, with carbon stocks estimated at 247 Gt C, including 193 Gt C aboveground and 54 Gt C belowground. The study used a combination of data from 4,079 in situ inventory plots and satellite Lidar and optical imagery to estimate carbon storage. The map provides methodologically comparable estimates for 75 developing countries, improving upon previous assessments that were often incomplete or outdated. The study highlights regional differences in carbon stocks, with Latin America accounting for 49%, sub-Saharan Africa for 25%, and Southeast Asia for 26%. The map also includes uncertainty estimates, with pixel-level uncertainties ranging from ±6% to ±53%, but constrained at larger scales to ±5% and ±1% for projects and national levels, respectively. The accuracy of the map was validated using ground data and error propagation through spatial modeling. The study emphasizes the importance of accurate carbon stock estimates for REDD initiatives, as deforestation and forest degradation contribute significantly to global greenhouse gas emissions. The benchmark map provides a spatially explicit and traceable approach to estimate forest carbon stocks, which is essential for monitoring and verifying carbon emissions reductions. The map also highlights the need for higher-resolution data and in situ sampling to reduce uncertainties at the pixel level. The study underscores the challenges in estimating forest carbon stocks, particularly in tropical regions, and the importance of integrating remote sensing data with ground-based measurements. The benchmark map serves as a critical tool for developing countries to assess their carbon stocks and support climate change mitigation efforts. It also aids in policy-making and project planning by providing consistent and spatially explicit data for national and regional assessments. The map is recommended for use at national and project scales due to the high uncertainty at the pixel level, and further improvements in data resolution and sampling are needed to enhance accuracy.A benchmark map of forest carbon stocks in tropical regions across three continents has been developed to provide a reliable estimate of biomass carbon stocks for REDD (Reducing Emissions from Deforestation and Forest Degradation) assessments. The map covers 2.5 billion hectares of forests in Latin America, sub-Saharan Africa, and Southeast Asia, with carbon stocks estimated at 247 Gt C, including 193 Gt C aboveground and 54 Gt C belowground. The study used a combination of data from 4,079 in situ inventory plots and satellite Lidar and optical imagery to estimate carbon storage. The map provides methodologically comparable estimates for 75 developing countries, improving upon previous assessments that were often incomplete or outdated. The study highlights regional differences in carbon stocks, with Latin America accounting for 49%, sub-Saharan Africa for 25%, and Southeast Asia for 26%. The map also includes uncertainty estimates, with pixel-level uncertainties ranging from ±6% to ±53%, but constrained at larger scales to ±5% and ±1% for projects and national levels, respectively. The accuracy of the map was validated using ground data and error propagation through spatial modeling. The study emphasizes the importance of accurate carbon stock estimates for REDD initiatives, as deforestation and forest degradation contribute significantly to global greenhouse gas emissions. The benchmark map provides a spatially explicit and traceable approach to estimate forest carbon stocks, which is essential for monitoring and verifying carbon emissions reductions. The map also highlights the need for higher-resolution data and in situ sampling to reduce uncertainties at the pixel level. The study underscores the challenges in estimating forest carbon stocks, particularly in tropical regions, and the importance of integrating remote sensing data with ground-based measurements. The benchmark map serves as a critical tool for developing countries to assess their carbon stocks and support climate change mitigation efforts. It also aids in policy-making and project planning by providing consistent and spatially explicit data for national and regional assessments. The map is recommended for use at national and project scales due to the high uncertainty at the pixel level, and further improvements in data resolution and sampling are needed to enhance accuracy.