Tropical forests are major reservoirs of biomass and play a critical role in regulating the global climate. Accurate mapping of aboveground biomass (AGB) is therefore essential for quantifying carbon stocks, monitoring forest dynamics, and supporting climate mitigation strategies. However, tropical forests exhibit notable spatial variation, resulting in increased spatial uncertainty when estimating AGB. Therefore, estimating AGB requires integrating multiple data sources. To provide new insights into AGB modeling in the biodiverse Western Ghats region of India, this study evaluates the integration of Sentinel-1 (S1) and Sentinel-2 (S2) data, leveraging the complementary strengths of radar and optical sensors. Three machine learning techniques, Random Forest (RF), Support Vector Machine, and Extreme Gradient Boosting, were compared for their performance in accurately predicting the AGB. RF was found to be the best-performing machine learning model, with an R 2 of 0.55, a root mean square error of 75.34 Mg/ha, and a mean absolute error of 57.79 Mg/ha. Texture parameters extracted from the multispectral bands of S2 and S1's backscatter emerged as the most important variables with a relative importance score of 34.9%. The spatially predicted AGB map from the best-performing RF model exhibited a range of 16.64 to 375.73 Mg/ha with a mean value of 215.37 Mg/ha, whereas the uncertainty value ranged from 11.78 to 121.01 Mg/ha. Based on the predicted map, a high AGB was observed in semievergreen forests, followed by the moist deciduous forests and tree plantations. Low AGB was recorded in dry deciduous forests. This study recommends the integration of S1 and S2 data along with several auxiliary datasets, including topography, climate, and canopy height, to substantially improve the accuracy of AGB estimation in tropical forests. The results of this study are promising and support reliable carbon stock assessments, thereby contributing to the sustainable management of these forests within the framework of the Reducing Emissions from Deforestation and Forest Degradation (REDD+) initiatives.