The present experiment investigates similarities in participants’ nonverbal and verbal behaviours when responding to baseline and investigative questions, comparing two different types of baselines. Police literature suggests to obtain a baseline through small talk, whereas academic literature underlines the importance of baseline and investigative themes to be comparable. First, a baseline was obtained (either small talk or comparable), then the investigative questioning started. During the investigative questioning, participants either truthfully reported a set of actions they had actually performed or lied about them. Findings revealed that truth tellers and liars in the small talk condition did not differ in their level of similarity when responding to the baseline and investigative questions. In the comparable truth condition, levels of verbal similarity between the baseline and investigative questions were higher for truth tellers than for liars, but only for one variable: spatial detail. Results therefore showed that a small talk baseline should not be used to assess interviewees’ credibility, and that a comparable truth baseline, although better than a small talk baseline, is still problematic.