Soil iron-bound organic carbon (Fe-OC) is essential for the long-term stability of the carbon pool. The Tibetan Plateau contains large soil carbon stocks and is experiencing rapid climate warming, yet the spatial variability and dominant controls on Fe-OC in alpine grasslands remain poorly constrained. Here, we conducted a comprehensive field survey, analyzing Fe-OC in 113 alpine grassland samples from the Tibetan Plateau and used machine learning to evaluate the spatial patterns and storage of Fe-OC in both topsoil (0-10 cm) and subsoil (20-30 cm). Our findings showed that Fe-OC content did not differ significantly between topsoil and subsoil, but varied among grassland types, being highest in alpine meadow (2.82 g/kg), followed by alpine steppe (1.95 g/kg) and desert steppe (1.27 g/kg). The proportion of Fe-OC to soil organic carbon (fFe-OC) did not differ significantly among grassland types, whereas mean fFe-OC was higher in subsoil (25.17%) than in topsoil (19.78%) overall. Random forest analysis showed that soil characteristics, particularly pH and cation exchange capacity (CEC), were more influential than climatic factors in explaining Fe-OC variation. Lower mean annual temperatures, lower pH, higher annual precipitation, and higher CEC promoted Fe-OC formation. Employing a robust model, we mapped the spatial distribution of Fe-OC across the Tibetan Plateau, observing decreasing concentrations with increasing latitude and higher levels in the southern regions. Conversely, fFe-OC concentrations increased with latitude, peaking in the northwestern areas. The estimated Fe-OC stocks of 0.657 Pg C in topsoil and 0.652 Pg C in subsoil of Tibetan Plateau alpine grasslands significantly enrich the alpine soil database, providing essential insights for carbon cycling in high-altitude ecosystems.