One of the major series of reviews in elementary and secondary education is the Best Evidence Encyclopedia, or the BEE. Up to now, findings for systematic reviews have largely been restricted to the reviews themselves, with few cases in which lessons learned across many reviews using similar methods can be synthesized. The completion of the Best Evidence Encyclopedia reading and math reviews permits a first opportunity to describe both substantive and methodological patterns across a broad set of studies involving all elementary and secondary grades, reviewed using a common set of review procedures. The purpose of the proposed paper is to synthesize both substantive and methodological findings across the five main Best Evidence Encyclopedia reviews of reading and math programs in grades K-12. The paper considers the following research questions: (1) Across subjects and grade levels, what effect sizes are associated with variations in a) textbooks, b) computer-assisted instruction, c) instructional process programs, and d) combinations of these? What subcategories within these types of interventions are associated with positive effects?; (2) How do summary outcomes of various types of programs vary across reading and math, and across elementary and secondary grades?; and (3) Across subjects and grade levels, how do effect sizes differ according to the following methodological criteria: (a) Use of random assignment; (b) Sample size; (c) Duration; and (d) Use of standardized measures. The "meta-findings" across the five Best Evidence Synthesis reviews suggest that strategies likely to improve student learning are those that improve the quality of daily instruction, increase students' active participation in the classroom, and help students learn metacognitive skills. Consistently successful programs, such as cooperative learning, teaching of metacognitive skills, and improved management and motivation approaches, as well as comprehensive programs such as "Success for All" and "Direct Instruction," all emphasize extensive professional development, typically including multi-day workshops, in-class followup, and clear guidance and extensive supportive materials for teachers. Technology can be effective to the degree that it also supports active instruction, cooperative learning, and improving classroom instruction. Changing curriculum or textbooks is rarely an effective strategy in itself, but may be an important element of comprehensive approaches that also incorporate instructional processes. Methodological patterns were also consistent across subjects and grade levels. Surprisingly, random assignment never made an important difference in effect sizes. Far more important were sample size, duration, and use of measures not inherent to treatments.