Empirical studies have found that people frequently lie in online dating profiles, and that the lower the online daters’ attractiveness, the more likely they are to enhance their profile photographs and lie about physical descriptors such as age, height, and weight (Toma and Hancock, 2010). This paper presents a mathematical model that shows how the above patterns of misrepresentation in online dating can emerge as a result of optimizing behavior in a two-stage game. We assume that online daters are distributed along an attractiveness index. In the first stage, online daters create a dating profile with the goal of landing the first meeting. In the second stage, a pair decides whether to have further future interaction. Online daters’ utility depends on the expected attractiveness of the partner with whom they have further future interaction. We show that in equilibrium, the amount of misrepresentation is inversely related to the online daters’ positions on the attractiveness index. To the best of our knowledge, no other study has presented a mathematical model of misrepresentation in online dating profiles.