Data are becoming increasingly important in science and society, and thus data literacy is a vital asset to students as they prepare for careers in and outside science, technology, engineering, and mathematics and go on to lead productive lives. In this paper, we discuss why the strongest learning experiences surrounding data literacy may arise when students are given opportunities to work with authentic data from scientific research. First, we explore the overlap between the fields of quantitative reasoning, data science, and data literacy, specifically focusing on how data literacy results from practicing quantitative reasoning and data science in the context of authentic data. Next, we identify and describe features that influence the complexity of authentic data sets (selection, curation, scope, size, and messiness) and implications for data-literacy instruction. Finally, we discuss areas for future research with the aim of identifying the impact that authentic data may have on student learning. These include defining desired learning outcomes surrounding data use in the classroom and identification of teaching best practices when using data in the classroom to develop students' data-literacy abilities.