The increased rainfall variability and extreme weather events in Somali Regional State led to longer periods of droughts, which directly affects availability and dependency on groundwater. Consequently, groundwater potential mapping has become a vital issue due to the rising demand for groundwater in the region. The main aim of this study is to assess groundwater potential in Jarar Administrative Zone using machine learning and remote sensing techniques. For this study, 19 groundwater conditioning factors (GWCFs) and 105 wells location data were used to develop and validate the random forest (RF), support vector machine (SVM) and Classification and regression tree (CART) models. Area under the curve (AUC) and five statistical indices for model performance evaluation were used to validate and compare the performance of each model. Based on the variance inflation factor analysis, 12 most important GWCFs for predicting groundwater potentiality (GWP) are selected, including lithology, rainfall, LULC, distance to stream landforms, FIS, soil texture, TWI, soil depth, drainage density, profile curvature and aspect. The model outputs showed that all the three models performed very well in assessing GWP (with AUC ≥ 0.7), despite the RF model (AUC = 0.958) outperformed both the SVM (AUC = 0.921) and CART (AUC = 0.897) models. The predicted groundwater potential maps also revealed that moderate (44.9– 48.7%) and high (30.9–37.4%) GWP classes are dominating the study area. The moderate to very high GWP classes in the Jarar Administrative Zone are generally found in the river system, areas with carbonate and colluvium dominated lithology that have greater infiltration rate, and in plain areas with high rainfall. At Woreda level, Degehamedo, Gunagado, southwestern parts of Degehabur and southeastern parts of Gashamo have relatively higher groundwater potential due to high rainfall and plain topography. Finally, the findings in this study may help policy makers, local managers and hydro-geologists better plan and manage groundwater resources in the region. Such findings could also help in the development of sustainable and long-term groundwater management strategies. Keywords: Groundwater potential mapping, Machine learning, Somali Regional State, Geospatial modeling