Land use significantly affects soil organic carbon (SOC) distribution and carbon dynamics. This study investigates the spatial variation of SOC across five soil depths under different land use types in Dali County, Guanzhong Plain. Field sampling and laboratory analysis were conducted, and both classical statistics and geostatistics were used to assess SOC patterns and spatial autocorrelation. Results showed that SOC content decreased with depth, averaging 7.47, 5.99, 4.51, 2.97 and 2.69 g/kg at 0-10, 10-20, 20-40, 40-60 and 60-100 cm, respectively. Strong spatial autocorrelation was found in the 10-20, 20-40 and 60-100 cm layers, and moderate autocorrelation in the 0-10 and 40-60 cm layers. SOC content increased from southeast to northwest, with higher values in the northern Loess Plateau. Total SOC stocks were highest in cultivated land and garden land (37.45 x 105 t and 42.66 x 105 t), primarily due to their larger area, while forest and grassland had lower total stocks (2.05 x 105 t and 1.74 x 105 t) despite comparable or higher SOC density in some layers. These findings demonstrate clear vertical and horizontal SOC distribution patterns shaped by land use and provide a scientific basis for carbon balance assessment and sustainable land management.