Paddies contain 78% higher organic carbon (C) stocks than adjacent upland soils, and iron (Fe) plaque formation on rice roots is one of the mechanisms that traps C. The process sequence, extent and global relevance of this C stabilization mechanism under oxic/anoxic conditions remains unclear. We quantified and localized the contribution of Fe plaque to C stabilization in a microoxic area (rice rhizosphere) and evaluated the role of this C trap toward global C sequestration in paddy soils. Visualization and localization of pH by imaging with planar optodes, enzyme activities by zymography, and root exudation by 14C imaging, as well as upscale modeling enabled linkage of three groups of rhizosphere processes that are responsible for C stabilization from the micro- (root) to the macro- (ecosystem) level. The 14C activity in soil (reflecting stabilization of rhizodeposits) with Fe2+ addition was 1.4−1.5 times higher than that in the control and phosphate addition soils. Perfect co-localization of the hotspots of β-glucosidase activity (by zymography) with exudation showed that labile C and high enzyme activities were localized within Fe plaques. Fe2+ addition to soil and its microbial oxidation to Fe3+ by radial oxygen release from rice roots increased Fe plaque (Fe3+) formation by 1.7−2.5 times. The C trapped by Fe plaque was 1.1 times higher after Fe2+ addition. Therefore, Fe plaque formed from amorphous and complex Fe on root surface act as a “rusty sink” for C. Upscaling by model revealed the global significance of C preservation within Fe3+ complexes in paddy soils. Considering the area of coverage of paddy soils globally, radial oxygen loss from roots and bacterial Fe oxidation may trap up to 130 Mg C in Fe plaques per rice season. This represents an important annual surplus of new and stable C to the existing C pool under long-term rice cropping.