Inventory, heuristic, statistical, and deterministic methods have been widely used in landslide susceptibility studies in recent years. This study aims to apply a GIS-based semi-quantitative approach (Analytical Hierarchy Process-AHP) to assess landslide susceptibility and determine landslide-prone areas at a regional level with medium-scale using publicly available datasets. The AHP was preferred due to its ability of correlating different parameters, which aids the researchers in producing relatively consistent landslide susceptibility maps. Three different major drainage basins in different geomorphological regions of Turkey that display different types of climate, different types of landforms and also, different seismic characteristics were selected. This is the first study ever where a landslide susceptibility assessment of the three unique major drainage basins of Turkey was performed. The sum of the areal extent of the three drainage basins exceeds 20% of Turkey's total footprint area, where the total footprint area of the landslide inventories in those regions is approximately 16% of the entire country's inventories. After each basin was analyzed through a 90 m spatial resolution with eight and ten causative factors, the results were validated, and susceptibility zoning was performed by utilizing a novel synthetic classification procedure developed for this study. The prime factors and weights governing landslides in each basin have resulted in different conclusions following the analyses. It was observed that a diverse range of the historical landslides occurred on the topographic wetness index (TWI) values of 12 and 13 derived from the 90 m resolution digital elevation model (DEM). During the study, a methodology was developed for performing semi-quantitative landslide zoning at a regional scale that may be used for assessing nationwide or continental landslide susceptibility, hazard, and risk analyses.