This paper reviews research on the achievement outcomes of mathematics programs for middle and high schools. There were 100 qualifying studies, 26 of which used random assignment to treatments. The findings revealed very small effect sizes for mathematics curricula and for computer-assisted instruction.