A strong base of research is needed to guide investments in teacher professional development (PD). This article considers the status of research on PD and articulates a particular direction for future work. Little is known about whether PD can have a positive impact on achievement when a program is delivered across a range of typical settings and when its delivery depends on multiple trainers. Despite a consensus in the literature on the features of effective PD, there is limited evidence on the specific features that make a difference for achievement. This article explains the benefits offered by experiments in addressing current research needs and--for those conducting and interpreting such studies--discusses the unique methodological issues encountered when experimental methods are applied to the study of PD. (Contains 18 notes.)