Background: There has been an abundant flow of information written about problem-based curricula in the basic sciences. However, there is little collective information about its use in the clinical setting. This article reviews the literature concerning the implementation of problem-based learning (PBL) in the clinical setting and draws conclusions regarding its misconceptions and feasibility. Summary: The authors found numerous articles that described PBL programs in undergraduate, graduate, and continuing medical education, but only a handful met the criteria outlined in Barrows' taxonomy of PBL. Many programs produced interesting innovations in clinical education, but they should not be labeled PBL. Conclusions: When writing about PBL, authors should describe how their program fits or departs from a PBL taxonomy and provide enough detail to permit others to replicate it. In addition, there remains a dire need for more experimental studies to warrant the extensive use of problem-based learning in the clinical setting.