An individual lies if they misreport honestly generated information. They cheat if they create fraudulent information. The field of behavioral ethics has sought to investigate both types of ethical decision making. To do so, scholars have deployed experiments that rely on incentivized self-report procedures in which the generation of information is often random. In this article, I demonstrate that incentivized self-report procedures are incapable of studying cheating behavior. This limitation, combined with the motivation to theorize about the nature of cheating, has caused the theoretical and empirical literatures to make competing predictions about the nature of deception. In this article, I introduce a new class of deception paradigms capable of identifying both cheating and lying behavior. Across four experiments, I demonstrate that cheating and lying are fundamentally different activities, each with their own set of environmental antecedent and behavioral consequences. By distinguishing between these types of deception, I reconcile several conflicting predictions presently asserted in the behavioral ethics literature.