In this article, the authors consider what can be learned from limited forms of evidence, for purposes of accountability and improvement of teacher education programs. They begin with a review of recent research on how evidence has been used to examine the effectiveness of teacher preparation and development. Using empirical evidence from a state with limited data capacity, they illustrate what can be learned from value-added measures as one form of evidence. As a case in point, the value-added scores for fifth-grade teachers are used to answer the question: To what extent are teachers' years of experience and the institutions from which they obtained their teacher training related to student achievement? The authors conclude with a discussion of the use of evidence by shifting the focus of accountability from simply responding to external requirements to developing internal practices that generate knowledge for improvement, and argue for collective responsibility among multiple stakeholders. (Contains 1 figure, 6 tables and 18 notes.)