Although most people are not better than chance in detecting deception, some groups of police professionals have demonstrated significant lie detection accuracy. One reason for this difference may be that the types of lies police are asked to judge in scientific experiments often do not represent the types of lies they see in their profession. Across 23 studies, involving 31 different police groups in eight countries, police officers tested with lie detection scenarios using high stakes lies (i.e., the lie was personally involving and/or resulted in substantial rewards or punishments for the liar) were significantly more accurate than law enforcement officials tested with low stakes lies. Face validity and construct validity of various lie scenarios are differentiated.