The simulated client method (SCM) has been used for over 20 years to study health care provider behavior in a first-hand way while minimizing observation bias. In developing countries, it has proven useful in the study of physicians, drug retailers, and family planning services. In SCM, research assistants with fictitious case scenarios (or with stable conditions or a genuine interest in the services) visit providers and request their assistance. Providers are not aware that these clients are involved in research. Simulated clients later report on the events of their visit and these data are analyzed. This paper reviews 23 developing country studies of physician, drug retail, and family planning services in order to draw conclusions about (1) the advantages and limitations of the methods; (2) considerations for design and implementation of a simulated client study; (3) validity and reliability; and (4) ethical concerns. Examples are also drawn from industrialized countries, related methodologies, and non-health fields to illustrate the issues surrounding SCM. Based on this review, we conclude that the information gathered through the use of simulated clients is unique and valuable for managers, intervention planners and evaluators, social scientist, regulators, and others. Areas that need to be explored in future work with this method include: ways to ensure data validity and reliability; research on additional types of providers and health care needs; and adaptation of the technique for routine use.