Diagnostic signs that may be useful as indicators of mendacity were identified, and their utility was demonstrated by examining the widely disseminated presentation of President Clinton's grand jury testimony of August 17, 1998. This testimony was examined for the presence of 23 clinically practical signs of dissimulation selected from the literature review. A segment of President Clinton's testimony that was subsequently found to be false was compared with a control period during the same testimony (internal control). A fund-raising speech by President Clinton to a sympathetic crowd served as a second control (external control). The frequencies of the 23 signs in the mendacious speech were compared with their frequencies during the control periods, and the differences were analyzed for statistical significance. No clinical examination was performed nor diagnosis assigned. During the mendacious speech, the subject markedly increased the frequency of 20 out of 23 signs compared with their frequency during the fund-raising control speech. President Clinton increased the frequency of 19 signs compared with their frequency during the control period of the same testimony. The 23 signs may be useful as indicators of the veracity of videotaped and scripted testimony. Verbal signs of lying identified include qualifiers/modifiers, expanded contractions, denials of lying, speech errors, pause fillers, and throat clearing. Nonverbal signs of lying include leaning and postural shifts, lip licking, lip puckering and tightening lips, drinking and swallowing, smiling and laughing, fewer hand gestures, and hand and shoulder shrugs. If these findings are confirmed through further testing, they could, with practice, be used by psychiatrists in assessing interviews. 3 tables and 31 references