Growing interest in teaching quality and accountability has focused attention on the need for rigorous studies and evaluations of professional development (PD) programs. However, the study of PD has been hampered by a lack of suitable instruments. The authors present data from the Teacher Knowledge Assessment System (TKAS), which was designed to administer Learning Mathematics for Teaching (LMT) measures. TKAS is being widely adopted in the evaluation of PD programs with over 500 separate program administrations and 16,000 teachers representing every major region in the country. The purpose of the current study was to leverage the TKAS dataset to develop a set of empirical benchmarks of effect sizes for designing rigorous studies of teacher professional development programs. The research design consisted of two goals: (1) to investigate the average effect sizes of over 259 PD programs in math, paying specific attention to variation across programs and variation across outcome measures and (2) to leverage the extensive professional development questionnaire describing the enacted professional development interventions and teachers using data collected from surveys of teachers and program providers. In conclusion, the authors report that, though their findings were preliminary, they provide potentially important guidance for the design of group randomized trials using teacher knowledge as an outcome measure. Most notable about the findings is the variation of effects between and across outcomes, suggesting that researchers should consider the specific math outcome that is most relevant for their intervention. Two tables and one figure are appended.