Abstract: Iron oxide in red soil is a critical factor regulating soil organic carbon sequestration. Our objective was to explore the relationship between soil organic carbon and iron oxide in uplands and paddies, which is beneficial for understanding the stabilization mechanism of soil organic carbon and provides scientific guidance for rational land use. Based on upland and paddy long-term fertilization experiments (over 35 years) in the red soil of southern China, the designed treatments included no fertilizer control (CK), chemical nitrogen (N), chemical nitrogen, phosphorus, and potassium fertilizers (NPK), and NPK combined with manure (NPKM). According to the method of Shavinov, the dry screening of soil aggregates was used to obtain large soil macroaggregates (>2 mm), small aggregates (0.25−2 mm), and microaggregates (<0.25 mm). All soil aggregates were used to determine soil organic carbon, soil dissolved organic carbon, complex iron oxide, free iron oxide, amorphous iron oxide, and iron activity. Compared with CK, NPK and NPKM treatments in uplands decreased soil macroaggregates but significantly increased soil small aggregates and microaggregates. N, NPK, and NPKM treatments in paddy reduced soil macroaggregates but increased small aggregates. The average organic carbon contents of soil aggregates were 8.21, 7.65, and 2.08 g·kg⁻¹ in paddy fields, and 2.93, 6.68, and 1.33 g·kg⁻¹ soil in uplands, respectively. The average contents of dissolved organic carbon in macroaggregates, small aggregates and microaggregates in paddy soils were 70.72, 79.83, and 30.29 mg·kg⁻¹, respectively, whereas those in upland soils were 7.27, 21.49, and 5.88 mg·kg⁻¹, respectively, under treatments of N, NPK and NPKM. For upland, the amorphous iron oxides in macroaggregates, small aggregates, and microaggregates under NPKM treatment were 2.45, 7.62, and 1.82 g·kg⁻¹, respectively, which was significantly higher than that in CK, N, and NPK. For paddy, the amorphous iron oxides in soil macroaggregates, small aggregates, and microaggregates under NPKM treatment were 5.27, 6.45, and 2.83 g·kg⁻¹, respectively. Compared with CK, NPKM treatment significantly increased the free iron oxides in each soil aggregate, and N treatment significantly increased only the free iron oxides in soil microaggregates. There was no significant difference in the free iron oxides in macroaggregates and small aggregates under N, NPK, and NPKM treatments. The iron oxides contents first increased and then decreased with soil aggregate size. For uplands, the amorphous iron oxide in small aggregates and microaggregates was positively correlated with soil organic carbon, with slopes of 0.64 and 0.45, respectively. The amorphous iron oxide in macroaggregates, small aggregates, and microaggregates was positively correlated with soil dissolved organic carbon, with slopes of 10.33, 7.36, and 7.34, respectively. For paddy, the free iron oxide in macroaggregates, small aggregates, and microaggregates showed a significant positive correlation with soil organic carbon, with slopes of 0.45, 0.29, and 0.84, respectively. The free iron oxide in soil microaggregates was positively correlated with soil dissolved organic carbon, with a slope of 23.12. There was a significant positive correlation between the content of amorphous iron oxide in small aggregates and microaggregates and soil organic carbon. The amorphous iron oxide in macroaggregates, small aggregates, and microaggregates was positively correlated with soil dissolved organic carbon, with slopes of 15.30, 17.91, and 13.78, respectively. In conclusion, the amorphous iron oxides has positive effect on soil carbon sequestration both in upland and paddy soils, while free iron oxides play an important role in soil carbon sequestration only in paddy fields.