Abstract: Based on the validated Penman-Monteith model, the evapotranspiration of Medicago sativa L., grown in the poplar shelterbelts in Maowusu Sandy area, Inner Mongolia, West China, was calculated. The results were as follows: (1) compared with the observational data of eddy-covariance method, the relative error of Penman-Monteith model was 14.73%(R²=0.84), which showed that Penman-Monteith model was reliable to estimate evapotranspiration of pasture in the shelterbelts; (2) evapotranspiration of alfalfa in different phonological periods was variable. Evapotranspirations at survival-branching stage, branching-budding stage, budding-flowering stage, and flowering-producing stage were 5.03, 71.10, 44.95 and 101.74 mm, respectively. The maximum of evapotranspiration was found at flowering-producing stage, accounting for 45.66% of the total amount. The spatial variation of alfalfa evapotranspiration was demonstrated with the pattern of "low- high- low". The further distance it is away from the shelterbelts, the more water alfalfa consumed. The maximum of evapotranspiration occurred at 115 m distance (about 7 times of the tree height) away from the shelterbelts with the amount of 234.41 mm; later on, evapotranspiration decreased gradually due to the impact of the other shelterbelt. Evapotranspirations of alfalfa in shelter belts and monoculture were 222.83 mm and 269.02 mm, respectively. By contrast with monoculture pasture, evapotranspiration of alfalfa in shelterbelts decreased by 17.2%. The biomass (dry weight) values of alfalfa in shelterbelts and by monoculture were 3287.28 kg/hm² and 2959.93 kg/hm², respectively, which showed that alfalfa in shelterbelts produced 11.1% biomass more than that by monoculture.