Respondent routers which allocate a single stream of respondents among multiple surveys are becoming more prevalent. Various methodologies exist for selecting a survey for a respondent in a router. Intuitively, it seems that combining multiple studies in a router has the potential to introduce a selection bias as opposed to studyspecific, invitation-based samples. However, this potential effect would be somewhat difficult to measure because the potential bias introduced at any given time will depend on the mix of qualification requirements for the studies in the router, as well as the rules used for routing between the various surveys. The research presented will introduce a method of calculating the potential bias introduced by any particular methodology of respondent selection within a router. In addition, the research also introduces a method of comparing the router bias between different router configurations using a simulation-based experimental design. Some common routing schemes will be compared under various conditions. This research will attempt to answer for each of the conditions examined if using a router as opposed to a study-specific invitation affects the results of the study, and if different routing approaches differ in the magnitude of their effect.