This paper is concerned with when, if ever, deceptive sex should be criminalised. It defends the idea that it is necessary to distinguish between deceptions that will generally be punishable from those that will not and puts forward a novel framework for carrying out this task. Based on the concept of identity nonrecognition, this framework also offers a new way of understanding what makes certain kinds of deceptive sex wrongful. After setting out this framework, I analyse each of the deceptions that is most often carried out within ‘ordinary’ contexts, explaining why only some of these should generally be punished. The paper concludes by suggesting that identity nonrecognition has the potential to inform criminalisation debates more generally, and that its relevance extends beyond discussions about deceptive sex.