Landslides are affecting a major portion of the mountainous region of the world and have a significant impact on socio-economic, infrastructure, livestock, and human elements. Several research has been carried out in recent years for the recognition, mapping, monitoring, and forecasting the landslide events in both small and large regions worldwide. Therefore, to understand the catastrophic phenomenon of landslides, remote sensing, and GIS-enabled statistical methods are widely used nowadays. Thus, landslide-caused calamities can be effectively managed and mitigated using a landslide susceptibility map (LSM) that identifies the potential landslide areas. The landslide susceptibility mapping using sampling ratio techniques is important for enhancing the accuracy of the LSM prepared with frequency ratio (FR)statistical models. Hence, this research evaluates the effect of sampling ratio on LSM for the Kangra district using the FR method. Therefore, to evaluate the efficiency of the sampling method, this work uses 559 landslide points obtained from the Bhukosh GSI portal. The obtained landslide points were then randomly divided into training and testing landslide points using four sampling ratios (60:40, 70:30, 80:20, and 90:10). With the combination of sampling ratios and seven causative factors including aspect, distance to road, distance to lineament, distance to river, elevation, lithology, and slope, the four LSM were prepared using the frequency ratio method. Further, the accuracy of the four resulting LSM was assessed using the area under the curve (AUC) receiver operator characteristics (ROC) curve. The results obtained for the 80:20 sample ratio was higher (figure 1a) (AUC = 0.892) than the other sampling ratios of 60:40(AUC = 0.872)(figure 1b), 70:30 (AUC = 0.879)(figure 1c), and 90:10 (AUC = 0.862)(figure 1d). Thus, the results showed that an 80:20 sampling ratio is effective and viable for identifying and mapping local landslides. Furthermore, the results showed that the sampling ratio needs to be explored for small-scale landslide mapping as well.