This study combines insights from the policy uncertainty literature and neo-institutional theory to examine corporate performance in implementing a government-sponsored voluntary environmental program (VEP) during 2004-2012 in Guangzhou, China. In this regulatory context, characterized by rapid policy changes, corporate performance in VEPs is affected by government surveillance, policy uncertainty, and peer pressures. Specifically, if VEP participants have experienced more government surveillance, they tend to perform better in program implementation. Such positive influence of government surveillance is particularly evident among those joining under high and low, rather than moderate uncertainty. Participants also perform better if they belong to an industry with more certified VEP firms, but worse if they are located in a regulatory jurisdiction with more certified VEP firms. At a moderate level of policy uncertainty, within-industry imitation is most likely to occur but within-jurisdiction imitation is least likely to occur.