Abstract <p> Landslides are a significant hazard in the Philippines, particularly on Mindoro Island, where diverse environmental and anthropogenic factors can contribute to slope instability. Physical assessments for landslides are both time-consuming and expensive, highlighting the need for rapid and cost-effective susceptibility mapping to enhance disaster preparedness and safeguard lives in vulnerable areas. This study aims to develop a landslide susceptibility map for Mindoro Island using fuzzy logic modeling. Fuzzy membership functions for landslide conditioning factors (i.e., annual rainfall, elevation, land use and land cover [LULC], slope, aspect, soil texture, topographic wetness index, and distance to faults, rivers, and roads) were derived from existing literature. Various aggregation methods were tested to combine fuzzy membership maps of the different factors to generate the final susceptibility map. The validity of the resulting maps was assessed using a remotely sensed landslide inventory map of the island employing different performance metrics. Results showed that landslide susceptibility in Mindoro varies spatially as influenced by different input variables and the choice of the aggregation method. The weighted aggregation method was the most accurate approach for combining fuzzy membership maps, providing a reliable and realistic susceptibility map. The sensitivity analysis revealed that the aspect, slope, LULC, and elevation are critical inputs for developing the final landslide susceptibility map. These findings underscore the influence of different environmental factors on landslide susceptibility modeling. Overall, this study demonstrates that a knowledge-based approach using fuzzy logic provides a simple, flexible, and computationally inexpensive framework with adequate performance for mapping landslide susceptibility.</p> Graphical abstract <p>The graphical abstract visually summarizes the methodological framework adopted for landslide susceptibility modeling on Mindoro Island, Philippines. The process begins with a geographic map showing the location and administrative boundaries of Mindoro, establishing the spatial context of the study. Susceptibility factors include topographic attributes (elevation, slope, aspect, and topographic wetness index) derived from digital elevation models; hydrological and infrastructural proximity variables (distance to rivers and roads); geological and land cover data (land use and land cover, distance to faults, and soil composition); and climatic data represented by rainfall patterns. These layers were resampled to 15 m spatial resolution to ensure compatibility and consistency throughout the modeling workflow. The central section of the graphical snapshot illustrates the application of fuzzy logic modeling, where each landslide conditioning factor was transformed into a fuzzy membership function to represent the likelihood of landslide occurrence. The right panel presents the resulting landslide susceptibility maps generated using various fuzzy aggregation methods - maximum, minimum, algebraic sum, algebraic product, average, weighted, and gamma functions. These maps depict the spatial distribution of different susceptibility zones, ranging from very low (green) to very high (red), based on the aggregation method applied. Meanwhile, the weighted map provides a more realistic delineation of susceptibility classes, supported by strong predictive performance, as reflected in key metrics such as accuracy, sensitivity, specificity, precision, and area under the curve (AUC). The inclusion of these indicators underscores the reliability of the model validation framework. At the bottom, the graphical abstract indicates the transition to sensitivity analysis, which quantifies the relative influence of each susceptibility factor on the performance of the landslide susceptibility model. The workflow highlights the flexibility and effectiveness of knowledge-based and fuzzy logic modeling as a practical approach for multi-criteria susceptibility analysis in landslide-prone areas of the model validation framework.</p>