For decades, there have been competing visions of how and where to educate students with disabilities (SWDs) in America’s K-12 schools. One conception is that general classrooms can accommodate the learning needs of virtually all children. A second approach calls for multiple placement options. Over the years, the context in which this disagreement has played out has changed as educators have shifted from a reliance on special classes to trust in general classes to enthusiasm for intensive instruction beyond the general class. Such variation in practice has influenced how researchers have explored relations between SWDs’ placement and their academic performance. Some of this research has been weak, producing unreliable findings. Some has generated more trustworthy results. All stakeholders would benefit from distinguishing the weaker studies from the stronger ones. Yet, to date, there has been an absence of such effort. In this paper, we provide a concise history of placement-achievement research and then review evidence spanning 50 years, bearing on how and where to educate SWDs. We conclude that the research on where to teach has generally been weak and inconclusive; the research on how to teach, stronger and more certain. Implications for educating SWDs are discussed.