The knowledge of soil electrical resistivity proves essential for a better earthing in order to ensure the protection of telecommunications and electrical energy networks. This study aims to estimate the

The knowledge of the ground electrical resistivity is essential to ensure the protection of electrical and telecommunications networks. However, the monitoring of its values is an expensive task which takes long time. Therefore, its prediction is important. This study investigates on predicting soil electrical resistivity using Artificial Neural Networks. Nine sites of our city (Lome, TOGO) were considered. After characterization of the resistivity data collected on these sites, two models have been developed: Multilayer Perceptron (MLP) and Radial Basis Function (RBF) networks. Relative Root Mean Square Error (RRMSE) and R2 (Linear Correlation Coefficient) have been used to evaluate each model performance. For the MLP model, the configuration [ABCDEF] is the most efficient with the RRMSE = 12.00%, R2 = 81.91% and 70 neurons under the hidden layer. For the RBF model, the configuration [BCDEF] is the most efficient with theRRMSE = 16.07%, R2 = 69.97% and 100 neurons under the hidden layer. In general, the results exhibit that the MLP outcome configuration [ABCDEF] is the most efficient with the best RRMSE = 16.07% and R2 = 69.97%. The letter A, B and C are the weather parameters and D, E, F are the geo-referenced coordinates of the measuring point. So far, research has not focused on predicting the electrical resistivity of the soil at a given location. Thus, the results of this study show that from meteorological data, it’s possible to predict this electrical resistivity.