Plantations are becoming more common globally as one of the important initiatives to mitigate global climate change, but the results on whether the soil organic carbon (SOC) can reach the level of natural forests are still inconsistent. Here, we conducted a meta-analysis of 418 paired observations, comparing plantations to adjacent natural forests (primary and secondary forests), from 47 published studies to explore the global patterns and associated drivers of SOC functional fractions (particulate OC, POC
mineral-associated OC, MAOC) and their ratios (the ratio of POC to MAOC, POC:MAOC
the ratio of POC to SOC, POC:SOC
the ratio of MAOC to SOC, MAOC:SOC). We found significant reductions of POC (42.4%, 35.9%), MAOC (19.4%, 15.2%), POC:MAOC (29.0%, 25.5%), and POC:SOC (18.2%, 18.9%) in plantations compared to primary and secondary forests. In contrast, MAOC:SOC in plantations had no change. The effects of plantations on POC, MAOC and their ratios were significantly affected by tree species, plantation age, soil type, and soil depth. Moreover, soil physical properties (soil bulk density, mean weight diameter), element contents (total phosphorus) and microbial communities (microbial biomass C) appeared to be drivers of lower POC, MAOC and their ratios in plantations. Our findings suggest that the reduction of SOC in plantations is mainly distributed in relative labile POC. The results reveal that the SOC functional fractions in plantations developed over time and were comparable to adjacent secondary forests after about 60 years, and plantations with native species will be more conducive to the formation of POC and MAOC. We emphasize that primary forests are not replaceable, and plantations with native species might be a reliable way for restoring a stable distribution of SOC.