Based on surface dead fuel data of Larix gmelinii (sunny slope) in Genhe City, mixed forest of Xylosma racemosum, Betula platyphylla and Betula dahurica (shady slope) in Oroqen Autonomous Banner, mixed forest of Betula platyphylla and Populus davidiana (sunny slope) in Yakeshi City and mixed forest of Betulaplatyphylla and Populus davidiana (shady slope) in Aershan City in great Xing’an Mountains forest and meteorological observation data during 2004-2019, the relationships between ground surface dead fuel moisture content and forest factors, meteorological factor inside and outside the forest were analyzed. The canopy density, surface dead fuel temperature and forest temperature had a negative correlation with ground surface dead fuel moisture, while the relative humidity had a positive correlation with ground surface dead fuel moisture content on sunny slope forest, and it was opposite on shady slope forest. The prediction models of ground surface dead fuel moisture content were established during the fire prevention period in different seasons by using stepwise regression method. The prediction model of sunny slope forest model was better than the shady slope model. The model was best in autumn, followed by spring and autumn, summer and it was worst during the total fire prevention period.