This article provides one of the first analyses of visuals in misinformation concerning COVID-19. A mixed-methods analysis of ninety-six examples of visuals in misinformation rated false or misleading by independent professional fact-checkers from the first three months of 2020 identifies and examines six frames and three distinct functions of visuals in pieces of misinformation: how visuals illustrate and selectively emphasize arguments and claims, purport to present evidence for claims, and impersonate supposedly authoritative sources for claims. Notably, visuals in more than half of the pieces of misinformation analyzed explicitly serve as evidence for false claims, most of which are mislabelled rather than manipulated. While this analysis uncovered a small number of manipulated visuals, all were produced using simple tools; there were no examples of “deepfakes” or other artificial intelligence-based techniques. In recognizing the diverse functions of visuals in misinformation and drawing on recent literature on scientific visualization, this article demonstrates the value in both attending to visual content in misinformation and expanding our focus beyond a concern with only the representational aspects and functions of misinformation.