Various audio-visual cues have been found to increase when people are deceiving. Similarly, these cues have also been reported to affect judgements of speaker certainty and confidence. In this study, both types of research have been combined in an experimental design based on feigned knowledge to determine the extent to which these cues can betray this type of deception. The first experiment in this study adopted a within-subject design where 16 participants took part in two question-answer sessions: one on their own degree topic and one on a topic that they did not study. In cases where the participant did not know the correct answer, they were required to deceive the interviewer by feigning knowledge. Participants then assigned a ‘feeling of knowing’ (FOK) score to each of their responses. Logit mixed-effects models revealed that facial expressions and gaze aversion were less likely to occur in both high FOK responses and the own-topic condition. Hand gestures also decreased with high FOK but adaptor behaviours were seen to increase. In a second experiment, a separate group of participants assigned ‘feeling of another’s knowing’ (FOAK) scores to audio-only and video-only responses from experiment one. The FOK scores assigned to these responses in experiment one significantly correlated with the FOAK scores although no difference in FOAK accuracy was found between the cue modalities. The results show a link between topic and response confidence: we are now able to talk about feigned knowledge in terms of speaker certainty and specific audiovisual cues in terms of occurrence likelihood. Cues have also been shown to have a communicative power that transcends from FOK to FOAK judgements, although which specific cues are the most revealing is yet to be pinpointed.