Problem-based learning (PBL) can take many different shapes but has as a common denominator that it builds on the principles of collaborative, constructive, contextual, and self-directed learning. Systematic review approaches that aim to provide insight in what features make PBL work generally fall short, as they tend to disregard the influential role of implementation contexts. The realist review approach seems to be promising in this respect, as this type of review aims to address the comprehensive question: What works for whom, in what circumstances, in what respects, and how? This article elaborates on the theoretical foundation of the realist review approach, provides examples and a step-by-step description of how to conduct a realist review, and sketches a promising perspective on the way in which realist reviews can contribute to furthering insight in PBL and its future development.