•Freeze-thaw induced landslides (FTILs) on grasslands were systematically examined.•Soil characteristics and topography were intrinsic factors controlling FTILs.•Increased rainfall and thickening active layer were direct drivers of FTILs.•Combining multiple monitoring methods is the trend for early warning of FTILs. Landslides induced by freeze–thaw processes on grasslands are one of the major geohazards, and their scale and frequency are increasing as the global warms. Freeze-thaw induced landslides degrade surface vegetation and soil properties, reduce biodiversity, intensify landscape fragmentation, and lead to losses in economy, human and animal lives. Despite substantial progress in research on landslides, there has been little study focused on how ground freeze–thaw events affect landslides. By critically analyzing previous studies, this paper proposes a conceptual framework for the forms and types, development, dominant factors, monitoring techniques, and impact mechanisms of freeze–thaw induced landslides. Landslides are controlled by soil characteristics and topographic slope, which are major intrinsic determinants. Increased rainfall, rising temperatures, and thickening active layer due to climate change are all direct drivers of freeze–thaw induced landslides. Vegetation conditions, animal behavior interference, and wind erosion all affect the occurrence and development process of landslides by modifying vegetation cover, soil physical and chemical properties, and structure. Currently, landslide monitoring techniques have evolved rapidly with improved efficiency and accuracy, but with only few applications for freeze–thaw induced landslides. There are a variety of prediction models for landslides, but few consider freeze–thaw effects and lack field validation. The new perspective on the occurring types and dominant factors enhances theoretical understanding of the formation mechanisms, which helps further monitor and analysis of freeze–thaw induced landslides. Future studies should concentrate on the coupling mechanism of multiple factors and the development of an accurate prediction system, which will greatly benefit the understanding and early detection of freeze–thaw induced landslides.