Recent studies have explored ways to increase cognitive load in liars to identify cues to deception. This study used a driving simulator as a load-inducing technique to explore differences between truth-tellers and liars during an investigative interview scenario and also investigated the effect of rehearsing lies in this context. Deception affected driving performance. Truth-tellers drove more slowly compared with their own baseline, whereas unrehearsed liars sped up. There was no difference in speed between truth-tellers and rehearsed liars. In addition, truth-tellers had significantly faster reaction times compared with their own baseline, than both rehearsed and unrehearsed liars. During the interviews, truth-tellers provided significantly more visual and auditory details and mentioned significantly fewer cognitive operations than liars. The findings add to the body of literature exploring the optimal relationship between cognitive load and secondary task performance to identify cues to deception.