Abstract: To identify the key drivers of forest fire occurrence in the West Sichuan Plateau and establish a scientific fire risk zoning system, multi-source data and constructed models were integrated to reveal the spatial distribution patterns of regional forest fires, providing scientific and technological support for precise fire prevention and ecological security barrier protection. Based on MOD14A1 satellite fire point data from 2003 to 2022 and 13 driving factors including meteorology, topography, vegetation, and human activity, a binary logistic regression model was constructed (model AUC=0.8, prediction accuracy 76.1%) for fire risk zoning. The results show that: (1) Seven factors significantly influence fire occurrence in the West Sichuan Plateau: distance to settlements, distance to roads, elevation, monthly average precipitation, monthly average land surface temperature, monthly average net shortwave solar radiation, and monthly average relative humidity. (2) The area proportions of low, medium, and high fire risk zones are 67.8%, 18.5%, and 13.4%, respectively, with high-risk areas concentrated in southern Ganzi Prefecture (Jiulong County, Kangding City, Yajiang County) and northern Aba Prefecture (Zoigê County). (3) Meteorological factors (low temperature, low humidity, low precipitation, and strong radiation) are the dominant drivers of forest fires. The combination with intensive human activity characteristics (within 6 km of settlements and roads) further exacerbates fire risk. (4) The fire risk zoning method based on driver quantification and spatial probability modeling can provide a scientific basis for dynamic monitoring and hierarchical management of forest fires in the Western Sichuan Plateau.