Wildfires are among the major drivers of land degradation and pose a serious threat to forest and rangeland ecosystems in arid and semi-arid regions by reducing vegetation cover, accelerating soil erosion, and diminishing ecosystem stability. This study aimed to model the spatial distribution of wildfire risk and assess the impacts of climate change on the spatial shift of fire-prone areas in the Central Zagros using the Maximum Entropy (MaxEnt) model. A total of 1,910 wildfire records collected between 2020 and 2024, together with 14 environmental variables including bioclimatic factors, topography, land use, and the Normalized Difference Vegetation Index (NDVI), were used to develop the model. After eliminating multicollinearity among predictor variables, the MaxEnt model was implemented, and the impacts of climate change were evaluated using the HadGEM2-CC and MRI-ESM2-0 General Circulation Models (GCMs) under the SSP245 and SSP585 scenarios for the years 2050 and 2070. The model demonstrated good predictive performance, with an AUC value of 0.82. Annual precipitation, land use, and slope were identified as the most influential predictors of wildfire occurrence, jointly accounting for 67.9% of the model contribution. Approximately 27.9% of the study area was classified as high wildfire risk, mainly concentrated in the western, northwestern, and central parts of the Central Zagros. Future climate projections indicated that climate change is likely to create new fire-prone areas across the region, with their extent ranging from 5.87% to 15.58%, depending on the climate model and emission scenario, highlighting a spatial shift in wildfire hotspots. The findings provide valuable information for identifying priority management areas, optimizing vegetation fuel management, implementing early warning systems, and planning proactive measures to reduce land degradation and support the sustainable management of arid and semi-arid ecosystems.