Theory-based evaluation (TBE) explores the how and why of program success or failure. Advocates of TBE claim that it produces information unavailable in traditional process and outcome studies. This article examines six published papers of TBEs. It finds that the authors of the papers do not always make explicit the relation of their data to the theory of the program. Nevertheless, it was evident in one or more cases that TBE identified unnecessary program components, located intermediary changes, raised new questions, contributed to a paradigm shift, highlighted the difficulties of taking successful pilot programs to scale, and provided clarity and focus for the evaluation. Interestingly, in none of the studies was the original theory completely right. Lessons for the future of TBE are drawn.