<p>This paper evaluates the impact of integrating various geological, geophysical, and geochemical datasets into mineral prospectivity mapping for undiscovered tungsten (W) and tin (Sn) deposits, using the Variscan French Massif Central (FMC) as a case study. Previous studies on the prospectivity of W–Sn deposits in Variscan terranes have primarily focused on geological and geochemical datasets, rarely integrating multi-method geophysical data. The FMC, located in the internal zone of the Western European Variscan belt, is a historic mining region known to host numerous granite-related W–Sn deposits. However, a comprehensive assessment is required to identify new prospective areas, including possible targets at depth. We applied a data-driven predictive approach using the disk-based association method coupled with random forest classification (DBA–RF) to predict the likelihood of undiscovered W–Sn deposits in the Puy-les-Vignes/Saint-Goussaud district (approximately 800 km²) located in the northern FMC. Favorable criteria from geological maps (lithostratigraphy, distance to faults, and known mineral occurrences) were analyzed, along with multiple datasets from airborne and ground geophysics (gravimetry, magnetics, electromagnetics, and gamma spectrometry), and stream sediment geochemistry (As, B, Be, Cu, Sn, and W concentration maps). The resulting prospectivity maps highlight several zones with a total area of 67 km² for W and 33 km² for Sn, which represent 8.3% and 4.1% of the study area, respectively. These zones coincide with known mineral occurrences, thereby validating the predictive model’s effectiveness. Additionally, new prospective zones extending well beyond (&gt;2&#xa0;km) the known mineral occurrences were identified, suggesting potential for discovery. This research demonstrates the value of using the DBA–RF method to integrate multiple geological, geophysical, and geochemical datasets in former mining districts where W–Sn occurrences are already known. The DBA–RF appears as a promising method for making the most of datasets acquired through historical and recent exploration surveys, and for enhancing mineral exploration programs by efficiently reducing the size of prospective areas. This methodological approach can serve as a robust framework in other prospective regions as a regional targeting tool to search for W–Sn deposits spatially associated with buried granitic intrusions.</p>