This review assembles what is known from studies and analyses of large-scale systemic reform initiatives aimed at mathematics and science education, especially those undertaken by state governments and the National Science Foundation. The review concentrates on qualitative investigations, which reveal whether and how these initiatives converge on the improvement of instruction. The evidence to date, though inherently incomplete, has much to say about the logic of systemic reforms, the avenues by which these reforms may reach the classroom, and the kinds of classroom-level impacts that are apparent so far, or are likely to occur in the near term. The implementation and effects of these reforms can be understood productively in the theoretical terms applied to the study of innovation and change, implementation of governmental policies, and professional and organizational learning. (Author)