The Concealed Information Test (CIT) aims to detect concealed information through differential physiological and behavioral responses to the concealed items. Although extensive experimental research has demonstrated the empirical validity of the CIT, the external validity of these studies has been questioned. One essential difference between experimental setups and realistic contexts is the voluntary act of committing the crime and concealing the critical (crime‐related) items. The present study examined whether the detection efficiency of the CIT alters under conditions of free choice to commit a mock crime. In the ‘choice’ condition, participants chose to perform a mock crime or a computerized task, while in the control, ‘instructed’ condition participants were instructed either to commit the mock crime or a computerized task. The results revealed no significant differences in the detection efficiency between the two conditions in electrodermal, respiration, and reaction time measures. Copyright © 2016 John Wiley & Sons, Ltd.