Practices of generating, analyzing, and using evidence play a central role in the Framework for K-12 Science Education and the NGSS. However, the construct of evidence remains largely underspecified in these documents, providing insufficient guidance on how to engage students with the broad and complex nature of evidentiary reasoning. This creates a risk of perfunctory and simplified implementation of evidence-based practices that misses the intent of the standards and does little to prepare students for reasoning with the complex, varied, and contentious evidence encountered in popular media or in advanced education. To address these challenges, we propose a theoretical framework, which we call "Grasp of Evidence," that complexifies the concept of evidence in ways that facilitate introducing more authentic forms of evidence and more sophisticated ways of engaging with evidence in science classrooms. Our approach focuses on promoting a "lay" grasp of evidence needed by competent outsiders as they engage with science in their everyday lives. The framework posits five dimensions. The first four dimensions capture what students should understand about how experts work with evidence: evidence analysis, evidence evaluation, evidence interpretation, and evidence integration. The fifth dimension focuses on how laypeople can use evidence reports themselves. Each of these dimensions of practice involves specific epistemic aims, epistemic ideals, and reliable epistemic processes for reasoning with and about evidence. We discuss these dimensions and their contribution to the conceptualization of evidence as well as provide some initial instructional implications and potential directions for future research.