In experiment 1, volunteer undergraduate students observed two films of crimes being committed. They were then asked to compose both a truthful and deceptive account of what they had observed in each of the films. The compositions were then videotaped as spoken by the volunteers. The one volunteer who was selected to be the "sender" for the experiment was the most convincing in her presentations. The sender presented the truthful and deceptive statements to 121 police officers and 146 psychology students, who were asked to assess the truthfulness of each statement. They were also asked to indicate whether their decision about truthfulness was made early, at the mid-point, or the final portion of the statement. The hypothesis that police officers would make their decisions early and that students would decide later was not supported. Further, deceptive statements tended to be judged accurately at the beginning of the statement and inaccurately at the end. The findings also indicated an initial lie bias for both students and officers. The second experiment varied the conditions of the first experiment by suppressing the sender's gestures and body movements that were visible in the first experiment. This experiment found that the sender's facial appearance did not reduce the lie bias found in the first experiment. Accuracy for detecting deceptive accounts decreased across time in both studies, and accuracy for truthful accounts increased only in the second experiment. The paper discusses how visual and verbal information contributed to these effects. 4 tables, 1 figure, 91 references, and appended supplementary data