High-stakes tests are the most heavily weighted measures in accountability systems developed in response to No Child Left Behind. While some studies show high-stakes accountability being related to test score gains, others suggest these policies do not improve achievement and often result in unintended consequences. To understand mechanisms driving positive and negative effects on student learning, it is necessary to look beyond achievement data and step inside classrooms. We present findings from 2 years of observation in 23 second and third grade classrooms, capitalizing on the combination of grade levels and seasons during which we observed. On the theoretically important domain of instructional support, we find classroom quality is lower when classrooms are under greatest pressure to increase test performance.