Karst aquifers are indispensable, yet vulnerable, resources; therefore, they require a comprehensive protection strategy. Since springs are the terminal points of the karst flow systems, knowledge of their distribution is a key element for the better understanding of groundwater flow, availability and vulnerability. The present study aims to introduce a data-driven analysis by the application of a spatial statistical technique (Weights of Evidence (WofE)) for the evaluation of factors influencing spring distribution in karst areas. A workflow was developed for investigating two questions: where will the springs locate, and where will the permanent springs evolve? This workflow has the potential for application to unconfined karst areas. This enhanced approach was applied to an unconfined transboundary aquifer, the Gömör–Torna Karst (HU and SK). The roles of five factors was statistically investigated: terrain elevation, distance to faults, distance of the carbonate–non-carbonate rock contact, distance to sinkholes, and precipitation distribution. The validation procedures confirmed the effectiveness of the approach. The resulting predictive maps are useful for decision-makers to delineate areas holding potential karst springs and to address water availability problems and protection measures. In addition, the WofE technique improved the comprehension of the geological conditions favourable for the formation of the springs.