Open AccessIn this article I present firstly the Norwegian training programme in investigative interviewing (K.R.E.A.T.I.V.), and secondly the concept of the cognitive interview (a method of interviewing witnesses and victims about what they remember from a crime scene). I argue that some of the techniques employed in both K.R.E.A.T.I.V. and the cognitive interview are problematic when communicating via an interpreter. In particular, this concerns the emphasis on the giving of a “free account”, which means that the interviewee is given the opportunity to talk without being interrupted. I also present the results of a survey that was sent out to Norwegian police officers. The purpose of the survey was to gain insight into what police officers found challenging when interviewing. Many of the respondents referred to different aspects of communicating via an interpreter. Finally, I discuss some possible solutions to these challenges.