This study presents a comprehensive landslide susceptibility assessment (LSA) for South Korea using a stacking ensemble of machine learning (ML) algorithms. The research aims to improve the accuracy and consistency of LSA by incorporating historical data and diverse natural factors, including past climate data. A total of 30 input variables were constructed, comprising topographical, environmental, and climatic factors. Sixteen ML algorithms were used as basic classifiers, and a stacking ensemble was applied to the four algorithms with the highest area under the curve (AUC). The CatBoost algorithm was found to be the most effective, with an AUC of approximately 0.89 for both assessments. Distance from roads, daily maximum precipitation, digital elevation model, and soil depth were identified as the most influential variables. The stacking ensemble enabled the creation of two landslide susceptibility maps, one for all landslide events and another for large-scale events (impact area ≥ 1 ha). The results highlight the importance of gradient boosting algorithms and the influence of various input variables on landslide susceptibility. The study concludes by providing a high-resolution (30 m) landslide susceptibility map for South Korea, which can be used for disaster prevention and mitigation.This study presents a comprehensive landslide susceptibility assessment (LSA) for South Korea using a stacking ensemble of machine learning (ML) algorithms. The research aims to improve the accuracy and consistency of LSA by incorporating historical data and diverse natural factors, including past climate data. A total of 30 input variables were constructed, comprising topographical, environmental, and climatic factors. Sixteen ML algorithms were used as basic classifiers, and a stacking ensemble was applied to the four algorithms with the highest area under the curve (AUC). The CatBoost algorithm was found to be the most effective, with an AUC of approximately 0.89 for both assessments. Distance from roads, daily maximum precipitation, digital elevation model, and soil depth were identified as the most influential variables. The stacking ensemble enabled the creation of two landslide susceptibility maps, one for all landslide events and another for large-scale events (impact area ≥ 1 ha). The results highlight the importance of gradient boosting algorithms and the influence of various input variables on landslide susceptibility. The study concludes by providing a high-resolution (30 m) landslide susceptibility map for South Korea, which can be used for disaster prevention and mitigation.