Climate change poses severe threats to mountain forest ecosystems, particularly in high elevation regions with complex topography. This study identifies and maps eco-climate refugia-areas of climatic stability and ecological persistence in the mountainous forests of Gilgit-Baltistan, Northern Pakistan, within the Himalayan-Karakoram- Hindukush Mountain system. These refugia serve as potential long-term habitats for forest species under changing climatic conditions. The research integrated multi-temporal geospatial datasets, including MODIS Land Surface Temperature (LST) and NDVI time series (2010-2025), 30 m SRTM/ASTER Digital Elevation Model, CHIRPS rainfall records, and forest cover data from Global Forest Watch and ESA World Cover. All datasets were preprocessed in Google Earth Engine and ArcGIS. Thermal stability was quantified using the interannual standard deviation of LST, while vegetation persistence was evaluated through NDVI trend and fluctuation analysis. Topographic buffering capacity was derived from elevation, slope, aspect, and Terrain Ruggedness Index (TRI). Rainfall stability was assessed via long-term precipitation variability. These layers were integrated using Multi- Criteria Suitability Modeling (MCSM) through weighted overlay analysis in ArcGIS Pro, with assigned weights: Thermal Stability (30%), Vegetation Persistence (25%), Elevation Buffering (20%), Rainfall Stability (15%), and Forest Density (10%). The study produced highresolution maps classifying the region into high-resilience refugia, moderate refugia, and vulnerable forest zones. Validation was performed through cross-comparison with existing protected areas and historical forest persistence data. This research provides the first comprehensive geospatial delineation of eco-climate refugia in Gilgit-Baltistan, offering valuable insights for climate adaptation planning, biodiversity conservation, protected area prioritization, and sustainable forest management in this climate-sensitive mountain region.