Abstract: Fine particulate matter (PM_2.5) has been paid more and more attention because of its negative effect on human health. Previous studies have shown that green space has an effect on PM_2.5 concentration, but it is not clear whether this effect correlates with the spatial distribution of PM_2.5 concentration. In addition, most of these studies focus on the change of PM_2.5 concentration in the interior of green space, and there is little research on the spillover effect of green space. Based on the 2018 remote sensing images and PM_2.5 raster data of urban districts in Hefei, the study area was divided into three regions, namely HH (High-High) effect area, LL (Low-Low) effect area and insignificant effect area through autocorrelation analysis of the spatial distribution of PM_2.5 concentration. According to the interpretation and selection of the green space, the research green spaces were obtained, PM_2.5 raster data were statistically analyzed by ArcGIS, and the correlation analysis and regression analysis were carried out by SPSS. The results showed that the spillover effect of green spaces in different regions had different effects on PM_2.5 concentration. In HH effect area, the average concentration of PM_2.5 decreased with the increase of the distance from the green spaces, but it increased in LL effect area. The effect of green space index on the variation of PM_2.5 concentration was different in different regions, NDVI_g had the greatest effect on PM_2.5 concentration in LL effect area, while S_g had the greatest effect on PM_2.5 concentration in HH effect area and insignificant effect area.