Even as the Internet and i-devices have changed so many parts of people's lives, educational testing has not changed substantially in decades. Though researchers and educators have for years raised ample concerns about existing tests, assessments have barely evolved. Further, despite years of expanded testing and greater school accountability, independent tests such as the National Assessment of Educational Progress and the Programme for International Student Assessment show that U.S. achievement has remained relatively flat. While one could blame many factors on stagnant achievement--including instruction, teacher quality, funding, or governance--assessment is equally under fire. In this article, the author looks beyond the current standardized tests to imagine what educational testing that increases learning could be. He examines: (1) the desired characteristics of high-quality assessments; (2) an assessment model that measures complex skills; (3) examples of better assessments, including a look at the forthcoming Race to the Top consortia assessment programs; and (4) recommendations for policy makers and practitioners on what it will take to get there. (Contains 2 figures.)