Abstract: Accurately evaluating the effect of farmland shelterbelt on crop yield plays an important role in shelterbelt construction and management at regional scale.A method evaluating the farmland shelterbelt′s effect on crop yield based on remote sensing(RS) and geographic information system(GIS) was put forward in the paper.First, by using multi temporal remote sensing images combined with the theory of management for shelterbelt, the managing phase of the shelterbelts was recognized, and the shelterbelts were divided into 4 stages, including mature, pre mature, post mature and updated stage.Besides, the growth status of shelterbelts was classified as good, average and bad according to the value of normalized differential vegetation index(NDVI).Therefore, the decision rule of protection effect for farmland shelterbelts was established on the basis of managing phase and growth status of shelterbelts.Then the spatial distribution map of protection effect of shelterbelts was obtained by combining a simple model and GIS.In order to analyze the farmland shelterbelt's effect on crop yield under similar natural conditions and remove other affecting factors, with the support from mechanism methodology and GIS, the crop potential productivity was computed according to the factors such as radiation, temperature, precipitation, and soil.The study area could be divided into 3 regions, i.e.low, middle and high production area.Besides, based on the meteorological data and the in situ measured data of crop yield, absorbed photosynthetically active radiation (APAR) inversed by remote sensing, light use efficiency, and harvest index, the crop yield was estimated at regional scale with the light use efficiency model.Finally, on the basis of above results, including spatial distribution of protection effect, crop potential productivity and estimated crop yield by remote sensing, we proposed a method evaluating the farmland shelterbelt′s effect on crop yield.Taking a sample area in shelterbelts key construction region of northern Changchun as the example, the farmland shelterbelt′s effect on maize yield in 2009 was evaluated.The spatial distributions of protection effect of farmland shelterbelts and crop potential productivity subregions were obtained in study area.Besides, we estimated the maize yield in 2009 with the MODIS NDVI and the sampled data of maize yield.The result showed the absolute error of estimated maize yield in study area was 467.75 kg/hm², and the relative error was 5.43%, which met the needs of accuracy completely for crop yield estimation by remote sensing.The change of maize yield was analyzed in the subregions with different protection effect of shelterbelts and crop potential productivity in study region, and the results showed that farmland shelterbelt could enhance maize yield to some extent.Especially, in the lower production area with poor natural condition, the farmland shelterbelt could significantly increase maize yield, and the yield increase rate could be up to 8.85%.The yield increase rate was 6.81% and 2.4% in middle and high production area, respectively.Consequently, under the better nature condition, the protection effect on crop yield could be weaken.In high production area, tree was predominant in competition and would affect crops negatively, which could led to the reduction of crop yield in some places.Therefore, the planning and construction of farmland shelterbelts should be in accordance with local conditions, so that farmland shelterbelts could maximize the benefits by using the smallest area.For poor natural condition and severe wind area, the construction of farmland shelterbelts should be strengthened.But in an area with better natural conditions, the separation distance of shelterbelts network should be increased to weaken the negative effect of shelterbelts on crop yield.The research can not only enrich the technological means in protection effect of farmland shelterbelt for spatial information technology, but also make the evaluation results for farmland shelterbelt′s effect on crop yield at regional scale more scientific and reasonable, and provide decision making support for the shelterbelt management.