We estimate monetary policy surprises (sentiment) from the perspective of three different textualsources: direct central bank communication (FOMC statements and press conferences), newsarticles, and Twitter posts during FOMC announcement days. Textual sentiment across sourcesis highly correlated, but there are times when news and Twitter sentiment substantially disagreewith the sentiment conveyed by the central bank. We find that sentiment estimated using newsarticles correlates better with daily U.S. Treasury yield changes than the sentiment extracted directlyfrom Fed communication, and better predicts revisions in economic forecasts and FOMCdecisions. Twitter sentiment is also useful, but slightly less so than news sentiment. These resultssuggest that news coverage and Tweets are not a simple echo chamber but they provide additionaluseful information. We use Sastry (2022)’s theoretical model to guide our empirical analysis andtest three mechanisms that can explain what drives monetary policy surprises extracted from differentsources: asymmetric information (central bank has better information than journalists andTweeters), journalists (and Tweeters) have erroneous beliefs about the monetary policy rule, andthe central bank and journalists (Tweeters) have different confidence in public information. Our empirical results suggest that the latter mechanism is the most likely mechanism.