Many states now possess the data and statistical methods that can produce teacher value-added scores and link them to preparation programs. It is important to understand the limitations of these measures and the inferences that they do and do not support. These limitations fall into three categories. First, value-added measures (VAM) provide information about only one of several important dimensions of teacher preparation program quality, focusing on one outcome measure, but not addressing other program characteristics, including the quality of program resources, the appropriateness of program content, and the contributions programs make to teacher learning. Second, comparing programs on the average VAM scores begs the question of whether mean performance is the most appropriate way to look at program quality. Third, the measurement of program graduates' VAM is strongly affected by the labor market for teachers, which weakens the inferences from VAM scores to the quality of preparation programs.