Low uptake of the COVID-19 vaccine in the US has been widely attributed to social media misinformation. To evaluate this claim, we introduce a framework combining lab experiments, crowdsourcing, and machine learning to estimate the causal effect of 13,206 vaccine-related URLs shared on Facebook on US vaccination intentions. We estimate the impact of misinformation flagged by fact-checkers was 50X less than that of content that was not flagged but nonetheless encouraged vaccine skepticism. Although misinformation reduced vaccination intentions significantly more when viewed, flagged content’s exposure on Facebook was limited. In contrast, unflagged stories highlighting rare deaths following vaccination were among Facebook’s most-viewed stories. Our work highlights the need to scrutinize factually accurate but potentially misleading content in addition to outright falsehoods.