<p>Spatial data integration is a more popular approach for regional-scale exploration of mineral deposits. The Kaiama area, located in the southern part of the Zuru Shist Belt, northwestern Nigeria, is known for its orogenic gold and cassiterite mineralization hosted primarily in metasedimentary rocks, pegmatitic granites, and shear zones influenced by Pan-African deformation events. However, the area remains underexplored due to limited systematic integration of geoscientific datasets for predictive mineral mapping. This study applies prediction-area (PA) analysis, weight sum modelling, multifractal analysis, and receiver operating characteristics/area under curve analysis (ROC/AUC) to delineate favourable zones for gold and cassiterite mineralization. The validity of spatial data, including satellite multispectral imagery, aeromagnetic, radiometric, and structural datasets, was tested using prediction area (PA) analysis. The weight sum modeling technique was employed for spatial data integration to develop mineral potential maps (MPMs) for gold and cassiterite. Discretization of the mineral potential maps (MPMs) for gold and cassiterite was achieved through multifractal analysis, while ROC/AUC analysis was utilized to evaluate predictive accuracies. Results from spatial integration using the weight sum model indicate that gold mineralization is most favorable in the southwestern, central, and northeastern parts of the study area, where mineral occurrences strongly correlate with geological structures and geophysical anomalies. Cassiterite mineralization favorability is highest in the southeastern and northern regions, primarily associated with pegmatitic intrusions and structurally controlled zones. The application of multifractal analysis suggests that the very high, high, and low potential classes for gold mineralization account for 13.46%, 25.8%, and 47.3% of the study area, respectively, while 13.25% represents the background value. Similarly, cassiterite mineralization favorability is characterized by 18.5%, 32.4%, and 38.1% for the very high, high, and low potential classes, with 10.8% representing the background value. The predictive accuracy assessment using ROC/AUC analysis revealed prediction accuracies of 81.2% for gold and 85% for cassiterite, confirming the effectiveness of the weight sum model as a reliable predictive tool for mineral exploration in the Kaiama area. This study highlights the significance of integrating geophysical, remote sensing, and statistical modeling techniques for effective mineral prospectivity mapping in underexplored regions.</p>