What we choose to assess in science is what will end up being the focus of instruction. US science standards once treated content and inquiry as fairly separate strands of science learning, with content standards stating what students should know and inquiry standards stating what they should be able to do. In its content coverage, these standards were also deemed to be a "mile wide and an inch deep." Assessments followed suit. By contrast, the National Research Council in its Framework for K-12 Science Education Standards presents a very different way of thinking about science proficiency. The framework articulates three interconnected dimensions of competence: disciplinary core ideas, crosscutting concepts, and scientific and engineering practices. The framework's three-part structure--practices, crosscutting concepts, core ideas--signals an important shift for science education. It also presents a challenge for the design of both instruction and assessment--finding a way to describe and capture students' developing competence along these intertwined dimensions. Thus it is critical that state education policymakers develop a vision for a balanced, coordinated system of assessments to promote effective science teaching and learning that takes into account both classroom assessments and large-scale monitoring assessments. Interim solutions will be needed that can simultaneously satisfy federally mandated testing requirements and allow space for change in classroom practice. It is worth noting that there is very limited evidence that accountability policies driven by large-scale assessment have led to improved student achievement. In contrast, the positive relationship between classroom assessment and student learning outcomes is well established. Assessment that closely aligns with curriculum and instruction and that engages students in the kinds of science learning described in the National Research Council's (NRC) Framework will return the focus to what is most important--the direct support of students' learning.