Objective To predict the environmental risk factors and spatial distribution of plague using geographic information system (GIS) and Maxent model.Methods Wulan County in Qinghai Province was chosen as the study area，45 positions of Himalaya marmot were processed by GIS software.The Himalaya marmot ecological environment variables including dem，slope，aspect，vegetation index，land surface temperature and land cover were extracted and analyzed.The distribution of Himalaya marmot prediction model was constructed based on Maxent model and ArcGIS software，and the environmental risk factors analysis was carried out.Results The area under the receiver operating curve (AUC) of the prediction result was 0.904，and the standard deviation was 0.077，indicating the predictive performance was good.The environmental risk factors analysis using Jackknife test indicated that the annual average normalized vegetation index，land cover and dem were the most important factors for the spatial distribution of Himalaya marmot.The percent contribution was 51.6%，21.7% and 12.4% in sequence.The spatial distribution prediction showed the region with color from blue to red with Himalaya marmot existence probability increased gradually，having more risks for the occurrence and transmission of plague.Conclusions The environmental risk factors and spatial distribution of plague foci are predicted successfully using geographic information technology and Maxent model.The result can provide important reference for different plague epidemics control and prevention.%目的 利用地理信息系统(GIS)和最大熵模型(Maxent)进行鼠疫的地理环境风险要素探测和空间分布预测.方法 选取青海省乌兰县为研究区，利用GIS软件空间化处理得到鼠疫疫源地内主要宿主动物喜马拉雅旱獭位置点45处;根据喜马拉雅旱獭的生境特点，提取与分析包括高程、坡度、坡向、植被指数、地表温度、土地覆盖等与喜马拉雅旱獭相关的多源地理环境变量;利用最大熵模型和ArcGIS软件构建旱獭的空间分布预测模型，并研究与鼠疫疫情相关的环境风险要素.结果 模型预测的受试者工作特征曲线下面积(AUC)平均值为0.904，标准偏差为0.077，模型总体精度良好;利用刀切法进行的环境风险要素分析表明，年均归一化植被指数、地表覆盖、高程对于喜马拉雅旱獭的空间分布影响最为重要，贡献率分别为51.6%、21.7%和12.4%;空间分布预测结果显示，颜色由蓝到红的区域喜马拉雅旱獭存在的可能性逐渐增大，鼠疫发生和传播风险也相应较大.结论 利用地理信息技术和最大熵模型可以进行鼠疫疫源地的环境风险要素探测和空间分布预测，研究结果可以为其他鼠疫自然疫源地的疫情防治和管理提供重要参考.