Purpose Despite high potential value for real‐life situations, detecting true and false intentions by groups of suspects have not been previously investigated. Method The experimental study had a set‐up in which participants (N = 232), half in dyads and half in quartets, planned for either a mock crime or a non‐criminal event. In structured individual interviews, all participants were asked one set of questions targeting their intentions (anticipated questions) and one set of questions targeting the planning phase of the intentions (unanticipated questions). We scored the level of detail and consistency in participants' interview responses. Results As predicted, questions on the planning phase were perceived as unanticipated and difficult to answer by both liars and truth tellers. Truth tellers' answers to the question on intent were perceived as more detailed compared to the liars. Cells of truth tellers and liars achieved an equally high within‐group consistency for their answers to the questions on the stated intentions, whereas cells of truth tellers achieved a higher within‐group consistency for the answers to the questions on the planning phase. Finally, truth tellers' descriptions of their intentions contained more information related to how to attain the stated goal, whereas liars gave more information related to why it was necessary to attain the stated goal. Conclusions Asking anticipated and unanticipated questions can be a successful way of eliciting cues to true and false intentions among small cells of suspects.