The performance of government microblogs varies substantially across regions，and it is intriguing to explain what drives the difference.In this paper we identify a group of variables that may affect government microblog from the supply-demand perspective.We use multisource data of Chinese prefecture-level cities to empirically test the hypotheses.The results reveal that government microblog is primarily driven by supply-side factors，while the effects of demand-side variables are largely limited.We find that government size，fiscal health，administrative rank，and internet penetration rate are positively related to government microblog performance.We discuss the theoretical contributions and policy implications of the results，and conclude with avenues for future research.