Two interviewers are often used by police and intelligence agencies, yet very little is known about the benefits of this approach or the best way to employ it. This thesis tested a two interviewer protocol to determine how the second interviewer should behave in order to elicit cues to deceit. Over four experiments, a second interviewer was employed (in various ways) in mock forensic interviews before examining the amount of detail and/or repetition in interviewees’ responses. The results offered four key findings: First, that a supportive second interviewer can elicit cues to deceit, notably by encouraging truth tellers to say more. Second, the impact of a second interviewer can be influenced by other factors, such as unexpected questions and the behaviour of both interviewers. Third, the results supported previous research by highlighting that consistency between two accounts is an unreliable cue to deceit. Finally, this thesis also suggests that deliberately mimicking interviewees may encourage truth tellers to provide more detailed and accurate responses. One explanation for the observed findings is that a supportive second interviewer provides a source of positive feedback, reinforcing and encouraging interviewees’ default response strategy. The results enabled a recommendation that forensic interviewers adopt a supportive demeanour. More research is needed before a supportive second interviewer should be introduced in real life interviews. Future research should seek to replicate the findings presented here and test a speaking second interviewer. While it was not clear that a supportive second interviewer generated rapport, overall, this thesis provides support for a rapport based approach to investigative interviewing.