Over many decades, land degradation, due to gully erosion and landsliding, has remained a serious threat to land-use management and economic sustainability in southeastern Nigeria. Several causative factors have been listed to be responsible for the continuous land degradation. In this study, multiple geostatistical and new indexical models were integrated to assess the role of soil engineering properties in the erosion of ten major gullies in southern Anambra State, southeast Nigeria. Gully soil samples were analyzed for properties such as particle size distribution, water content, Atterberg limits, compactness, bulk density, specific gravity, shear strength, porosity and permeability coefficients. The geotechnical analyses revealed that the gully materials are weak and have poor erosion-resistant features. The results revealed that the soils' engineering properties influence their erodibility. Interparameter relationships observed from correlation, factor, and cluster analyses showed that the engineering parameters jointly impact the erosion processes. Novel indexical methods, sand-gravel ratio (SGR) and sand-fines ratio (SFR), alongside erodibility factor confirmed that the soils are erodible. A novel entropized soil loss rate (A E ) model estimated high cumulative annual soil loss rate for the area at 0.9705. Overall, it was revealed that SW, SC, and SP soils are far more susceptible to gully erosion than CL and GP soils in the area. Artificial neural network models were developed and used to accurately (with low model errors and R 2 range of 0.956-0.997) predict the SGR, SFR and A E . The findings of this research would be helpful in advancing soil erosion investigations and in planning for effective mitigation measures, with respect to engineering properties.