Introduction: Ischemic stroke results in activation of microglia, which may polarize towards a pro-inflammatory (M1) phenotype and/or an anti-inflammatory (M2) phenotype. Enhancer of zeste homolog (EZH) 2 is a histone-lysine N-methyltransferase enzyme. Accumulating evidence has suggested EZH2 is key modulator of microglia polarization by epigenetically regulating gene expression. We here investigated whether microglial-specific deletion of EZH2 leads to a beneficial protective effect in stroke in vivo . We further tested the therapeutical potential of an EZH2 inhibitor after stroke. Methods: Aged male mice were subjected to 60-minutes middle cerebral artery stroke. Tamoxifen administration was started 30 days prior to stroke to induce genetic deletion of microglial EZH2 in CX3CR1-creER/EZH2-floxed mice. EZH2 floxed mice were used as controls. EZH2 inhibitor GSK343 was I.P. injected (once a day for two consecutive days), starting three hours after stroke onset. Mice were sacrificed for immunohistochemistry and crystal violet staining (brain infarct assay) after behavior tests at 3 days after stroke. Results: The expression of microglial EZH2 was significantly abrogated in KO mice compared to the control floxed mice (144±15.43 vs. 50.65±4.99 cells/mm 2 , N=5/each group, P<0.01). EZH2 deletion reduced brain infarct volume (29.27±2.23% vs. 6.07±0.88%, N=7/each group, P<0.001) and improved functional outcome assayed by adhesive removal test (59±13.1 sec, N=7 in floxed control vs. 26.28±4.1 sec, N=12 in KO, P<0.01). Mechanistically, microglial EZH2 deletion led to a decrease in expression of M1 marker iNOS (170±14.78 vs. 76.65±11.38 cells/mm 2 , N=4/each group, P<0.05), an increase in M2 marker Arg1 (96.64±11.48 vs. 203.3±22.02 cells/mm 2 , N=4/each group, P<0.05) co-stained in microglia (Iba1). Finally, GSK343 treatment robustly reduced infarct volume (37.1±2.97% vs. 21.9±2.92%) and increased latency to fall in hang-wire test (23.5±3.5 vs. 45.9±7.1 sec) (N=8/group, p<0.05) 3 days after stroke. Conclusions: Both genetic deletion of EZH2 in microglia and pharmacological inhibition of EZH2 improved stroke outcome in aged. The effect may be due to limiting microglial M1 polarization and enhancing M2 polarization.