Dishonest responding is an important part of the behavioral repertoire and perfectly integrated in communication and daily actions. Thus, previous research aimed at uncovering the cognitive mechanisms underlying dishonest responding by studying its immediate behavioral effects. A comprehensive account of the aftereffects of this type of behavior has not been presented to date, however. Based on the methods and theories from research on task switching, we, therefore, explored the notion of honest and dishonest responding as two distinct intentional sets. In four experiments, participants responded either honestly or dishonestly to simple yes/no questions. Crucially, robust switch costs were found between honest and dishonest responding when questions succeeded promptly (Exp. 1) but also when an unrelated task intervened between questions (Exp. 2). Surprisingly, responding dishonestly to a question also affected responses in the subsequent intervening task in terms of a more liberal response criterion. Time to prepare for the upcoming intentional set further induced asymmetrical switch costs (Exp. 3). Finally, a novel control condition (Exp. 4) allowed us to pinpoint most of the observed effects to negation processing as an inherent mechanism of dishonesty. The experiments shed new light on the cognitive mechanisms underlying dishonesty by providing strong support for the concept of distinct mental sets for honest and dishonest responding. The experiments further reveal that these mental sets are notably stable and are not disturbed by intervening task performance. The observed aftereffects of dishonest responding might also provide a potent extension to applied protocols for lie detection.