AbstractSoil organic carbon (SOC) decomposition is inherently sensitive to temperature. As such, a big concerning is the potential SOC loss under climatic warming, but field empirical evidences are lacking, particularly in tropical forest soils in which ∼10% of global SOC is stored. Recently Nottingham et al. (2019) assessed the data collected from a novel experiment translocating soils across a 3000 m tropical forest elevation gradient to mimic temperature changesin situ, and concluded that warming caused considerable SOC loss. However, this conclusion was based on a metric with a strong assumption that soil cores translocated to other elevations on average had the same initial SOC content to control soil cores reinstalled at their original elevation. Because of limited replicates (n=3) in the data, an approach ignoring spatial heterogeneity of SOC content may undermine the credibility of the results. Here, we used a nonparametric bootstrap approach to re-analyze the data, explicitly taking data variability into account. Contrary to Nottingham et al. (2019), we found that SOC content did not show significant differences among translocated soils from the same elevation origin. Further looking into six chemical fractions determined by13C NMR spectroscopy shown that they had similar, insignificant response to translocation-induced temperature changes, which also does not support the conclusion of Nottingham et al (2019) that labile SOC is more sensitive to warming. We concluded that temperature changes did not significantly alter either total SOC content or its six chemical fractions after five years of shift of temperature regimes in tropical forests. This may largely due to thermal adaptation of microbial decomposition and environmental constrains (e.g., low pH) which suppress the effect of temperature changes. Longer term experiment with more sampling replicates are required to maximize the value of soil translocation experiments to address the effect of warming on SOC dynamics.