From April to September in 2019, the soil evaporation of Camellia oleifera forest in hilly area of Changsha was measured by micro lysimeter, and the local evapotranspiration of forest was monitored by ETgage simulated evapotranspiration sensor. The temporal and spatial changes of soil evaporation in hilly area of Changsha were compared and analyzed, and the relationship between soil evaporation and meteorological factors and soil water content was analyzed. The results showed that the total soil evaporation of Camellia oleifera forest in the hilly area of Changsha was about 192.15 mm, with a daily average of 1.05 mm; in different spatial positions, the soil evaporation on the sunny, absolute and relative shady surfaces decreased in turn, with an average of 1.14, 0.98 and 0.94 mm/d, respectively. In the early stage(April to May), the influence of air temperature, ground temperature, wind speed, solar radiation intensity and air relative humidity on soil evaporation decreased in turn; in the middle stage(June to July), the influence of air temperature, air relative humidity, solar radiation intensity, ground temperature and wind speed on soil evaporation decreased in turn; in the later stage(August to September), the influence of ground temperature, solar radiation intensity, air temperature, wind speed and air relative humidity on soil evaporation decreased in turn. The results showed that the correlation between soil evaporation and soil water content was small in the early stage, and there was a significant positive correlation between soil evaporation and soil water content in 5 cm soil layer in the middle and late stage, and the soil evaporation front mainly occurred in 0-5 cm soil layer; the local evapotranspiration of Camellia oleifera forest land in Changsha hilly area was about 2.12 mm/d, and the local transpiration of Camellia oleifera tree was about 1.11 mm/d. The trend of local evapotranspiration measured by ETgage was similar to that measured by micro lysimeter, with a significant positive correlation between the two methods.