Background: Computer-mediated communication (CMC) is an important part of work life. However, this communication can be dishonest, and when people attempt to judge dishonesty, irrespective of the cues available, they tend to rely on a few nonverbal cues that are not the most reliable. Literature review: According to leakage theory, CMC modes differ from each other in the number of cues to deception they can transmit, potentially affecting one's ability to detect deception in a given medium. There is considerable research on peoples’ use of nonverbal cues across CMC modes to evaluate deception, but limited understanding of the choices they make and the extent to which their deception judgments are impaired or helped by cues they have access to for different CMC modes. Research questions: 1. To what extent are the nonverbal cues that people say they rely on to detect deception shaped by the medium that they use for communication? 2. What are the effects of nonverbal cue availability on deception detection success? Methodology: We conducted an experiment with 132 veracity judges from New Zealand and Jamaica, who observed interview segments in Spanish and Hindi (languages that they did not understand) to isolate the effects of nonverbal cues. They determined the veracity of each segment and listed the things that guided their judgment. Results/discussion: The results suggest that when certain nonverbal cues are available, such as gaze aversion, these suppress attention to more reliable cues (e.g., voice pitch) when judging deception. Redirecting attention to more reliable cues is therefore important. Unexpectedly, cue choice also varied across language by medium. Conclusions: The findings extend the understanding of people's use of nonverbal cues and the extent to which certain cues distract in the deception judgment. Although people rely on vocalic cues in audio-only media and kinesic cues in video-only media, they tend to rely mostly on, and are distracted by, a few kinesic cues for full audiovisual media, even though vocalic cues are available. We also found that people can successfully detect cues to deception, even when their communication mode is relatively bereft of useful information. However, the availability (or lack) of nonverbal cues was not a factor in deception detection success. To improve detection, deception training that targets reliable cues for different media is needed.