Abstract: formation mechanism of environmental regulation on PM _2.5 pollution by using the panel grey correlation, spatial measurement and mediation effect model. The results are shown as follows:（1) There exists significantly negative spatio-temporal correlation between environmental regulation and PM _2.5 pollution. The mean value of the temporal correlation intensity is negative, presenting a fluctuation of “W” pattern. The negatively spatial correlated cities account for a relatively high proportion.（2) Environmental regulation has an obvious inhibitory effect on PM _2.5 pollution, while the spillover effect of environmental regulation in neighboring cities has not played a positive role in promoting the reduction of local PM _2.5 pollution. In terms of the control variables, urbanization and PM _2.5 pollution show “inverted U-shaped” changes relationship, the size of the city, the level of foreign capital utilization, and the green coverage of the built-up area have a significantly positive impact on PM _2.5 pollution, and the level of informatization, advanced industrial structure, and energy intensity have a significantly negative impact. The impact of capital fails the significance test.（3) The important mediating variables of environmental regulation to reduce PM _2.5 pollution are industrial ecological efficiency and technological progress level respectively. The indirect effect of environmental regulation to reduce PM _2.5 pollution by improving industrial ecological efficiency accounts for 55.69%, while the indirect effect of environmental regulation to reduce PM _2.5 pollution by improving technological progress level accounts for 4.85%.（4) Environmental regulation exhibits spatial heterogeneity effect on PM _2.5 pollution, which cannot effectively reduce PM _2.5 pollution of cities in eastern, central, and northeastern regions, but can significantly lessen the PM _2.5 pollution in western cities.