AbstractCorrective or refutational posts from ordinary users on social media have the potential to improve the online information ecosystem. While initial evidence of these social corrections is promising, a better understanding of the effects across different topics, formats, and audiences is needed. In three pre-registered experiments (N = 1944 UK, N = 2467 Italy, N = 2210 Germany) where respondents completed a social media post assessment task with false and true news posts on various topics (e.g., health, climate change, technology), we find that social corrections reduce perceived accuracy of and engagement with false news posts. We also find that social corrections that flag true news as false decrease perceived accuracy of and engagement with true news posts. We did not find evidence to support moderation of these effects by correction strength, anti-expert sentiments, cognitive reflection capacities, or susceptibility to social influence. While social corrections can be effective for false news, they may also undermine belief in true news.