We review the use of real-effort tasks in economic experiments. To this point, the paradigm has been mostly used to model principal–agent relationships in the laboratory, the focus of our review. We first discuss the rationales for choosing between real and chosen effort when designing an experiment. To facilitate this discussion, we present a taxonomy of the common tasks that people have used, discuss some issues to keep in mind when implementing a real effort task and then survey the limited literature that compares the two methods. We end by offering a few recommendations on topics that could use additional investigation.