This study integrates machine learning (ML) ensembles to map landslide susceptibility in northern Pakistan, an area prone to landslides due to rugged topography, frequent seismic activity, and seasonal rainfall. The research employs a dataset of 228 landslide inventory maps and uses a random forest classifier and correlation-based feature selection (CFS) to identify the twelve most significant parameters influencing landslides. The parameters include slope angle, elevation, aspect, geological features, and proximity to faults, roads, and streams. Slope is identified as the primary factor, followed by aspect and rainfall. The models, validated with metrics such as AUC, ACC, and Kappa, show that logistic regression (LR) outperforms baseline models, while XGBoost excels among ensemble algorithms. The study highlights the practical effectiveness of these models in developing precise landslide susceptibility maps, which can guide risk-mitigation strategies and policies in geohazard-prone regions. The results also emphasize the importance of integrating ML methods to enhance the accuracy and reliability of landslide susceptibility assessments.This study integrates machine learning (ML) ensembles to map landslide susceptibility in northern Pakistan, an area prone to landslides due to rugged topography, frequent seismic activity, and seasonal rainfall. The research employs a dataset of 228 landslide inventory maps and uses a random forest classifier and correlation-based feature selection (CFS) to identify the twelve most significant parameters influencing landslides. The parameters include slope angle, elevation, aspect, geological features, and proximity to faults, roads, and streams. Slope is identified as the primary factor, followed by aspect and rainfall. The models, validated with metrics such as AUC, ACC, and Kappa, show that logistic regression (LR) outperforms baseline models, while XGBoost excels among ensemble algorithms. The study highlights the practical effectiveness of these models in developing precise landslide susceptibility maps, which can guide risk-mitigation strategies and policies in geohazard-prone regions. The results also emphasize the importance of integrating ML methods to enhance the accuracy and reliability of landslide susceptibility assessments.