Previous studies on upland soils showed that 13C natural abundance can successfully reveal C stabilization pathways between aggregates and soil organic matter (SOM) density fractions. The direction of C stabilization in paddies can, however, deviate from that in upland soils owing to i) periodic drying–rewetting cycles, with oxygen pulses under oxic conditions, and thus, shifts in microbial processing of organic residues, and ii) intensive organic and mineral fertilization. To trace C stabilization in paddies, soil was sampled from a long-term field experiment under an unfertilized Control and NPK, NPK + straw, and NPK + manure fertilizer regimes. Soil was analyzed for total C, microbial biomass (MB), and dissolved organic C, and separated into three classes based on aggregate size (>250 μm, 53–250 μm, and <53 μm) followed by density fractionation of each class to obtain free and occluded light fractions as well as dense and mineral heavy fractions. Pathways of C were determined based on C content and δ13C in all pools. The highest increase in total C (69%) was in NPK + manure, whereas the MBC increased with fertilization by at least 29% compared with the Control. All fertilizers increased macro-aggregation by at least 111% compared with the Control. The highest C content in the aggregates was in the mineral fractions of macroaggregates. Fertilization decreased the δ13C of total SOM compared with the Control, indicating suppressed decomposition of organic compounds. Aggregate size classes showed a typical δ13C enrichment trend from macro-to microaggregates, reflecting similarities between paddy and upland soils. A detailed scheme of C flows within aggregates and SOM fractions based on the δ13C natural abundance revealed the following general sequence: mineral → dense → free light → occluded light fractions. This trend, which is partly opposite to that observed in upland soils, reflects the anoxic and variable redox conditions of paddies. This facilitates the predominant stabilization of recent C input in the mineral fraction with Fe oxides due to Fe2+/Fe3+ dynamics, whereas light fractions are processed by microorganisms mainly in periods without overflooding. The C pathways in the two heavy fractions were separate from those in the two light fractions, which also indicates differences in the C stabilization processes between paddy and upland soils. Thus, the present study provides further detailed insights into the C stabilizing mechanisms in paddy soils which depend on management. © 2020 Elsevier Ltd