Commentaries regarding appropriate methods for researching professional development have been a frequent topic in recent issues of "Educational Researcher" as well as other venues. In this article, the authors extend this discussion by observing that randomized trials of specific professional development programs have not enhanced our knowledge of effective program characteristics, leaving practitioners without guidance with regard to best practices. In response, the authors propose that scholars should execute more rigorous comparisons of professional development designs at the "initial stages" of program development and use information derived from these studies to build a professional knowledge base. The authors illustrate with examples of both a proposed study and reviews of evidence on key questions in the literature.