Despite increasing expectations that researchers and funding agencies release their data for reuse, concerns about data misuse hinder the open sharing of data. The COVID‐19 crisis brought urgency to these concerns, yet we are currently missing a theoretical framework to understand, prevent, and respond to research data misuse. In the article, we emphasize the challenge of defining misuse broadly and identify various forms that misuse can take, including methodological mistakes, unauthorized reuse, and intentional misrepresentation. We pay particular attention to underscoring the complexity of defining misuse, considering different epistemological perspectives and the evolving nature of scientific methodologies. We propose a theoretical framework grounded in the critical analysis of interdisciplinary literature on the topic of misusing research data, identifying similarities and differences in how data misuse is defined across a variety of fields, and propose a working definition of what it means “to misuse” research data. Finally, we speculate about possible curatorial interventions that data intermediaries can adopt to prevent or respond to instances of misuse.