In this article we briefly examine the meta-analysis of systematic phonics instruction published in "Teaching Children to Read," the influential 2000 report by the National Reading Panel. A brief review of this study, and the ensuing reanalysis by Camilli, Vargas, and Yurecko, is given. Following this, we report new analyses that substantially alter previous interpretations of the effect of systematic phonics instruction. In the second part of the article, we examine how "knowledge" created through meta-analysis has entered the domain of public discussion. We conclude by considering how research studies consistent with the intent of the No Child Left Behind law can be designed to engender more effective reading instruction.