Assessment Criteria Indicative of deception (ACID) is a statement analysis technique that combines careful interviewing designed to facilitate the detection of deception with empirically-derived content criteria related to both the phenomenal and objective nature of deception. The current study of ACID had three goals. The first goal was to replicate ACID with a population of incarcerated witnesses. The second goal was to study the extension of this procedure by adding coherence and the type-token ratio for dependent measures. The third goal was to investigate the possibility of gender differences in verbal interpersonal deception. In meeting the first and second goals, this study demonstrated honest statements were longer and more detailed than deceptive statements, and that honest statements specifically had more words and more unique details added as a result of the recall enhancement effect of the interview. Honest statements were more coherent and had lower type-token ratios. In meeting the third goal, this study demonstrated that women's statements were more coherent and contained more words than male statements. Deceptive females gave less detail, fewer word, and had higher type-token ratios than deceptive men. Honest women gave more words and more detail than honest men. Importantly, coherence is very dependent upon gender, so that an honest man is more likely to tell an incoherent story than is a deceptive woman. Using gender-specific norms significantly increased the classification accuracy of a discriminant function analysis. Overall, 74 of 83 participants were accurately classified as honest or deceptive using the dependent measures. Future research is necessary to firmly establish the presence and cause of gender differences in verbal deception.