Soil organic carbon plays a vital role in the arid and semiarid regions. This study aimed to predict and map soil organic carbon content at soil depth intervals of 0-0.3 m, 0.3-0.6 m, 0.6-0.9 m, and 0.9-1.2 m in alluvium soils along Blue Nile and River Nile, Sudan. Ordinary kriging (OK) technique was used as a geostatistical tool and applied to model the spatial variability of soil organic carbon in the study area. A total of 38 soil profiles were excavated in the study area and 152 samples from the four depths intervals were collected for determining organic carbon content. Results revealed that, the spatial autocorrelation for the different soil layers was moderate to weak with a nugget to sill ratios ranging from 0.21 to 0.86 suggesting their controlled by both intrinsic and extrinsic factors. The root mean square error standardized (RMSE) of the predictions ranging from 0.79 to 0.83 indicating that the model which generated by ordinary kriging was correctly estimating the variability of soil organic carbon in the study area.