This paper provides an ethnographic account of understandings of ‘ethnic’ difference in Khovd province, Mongolia and attempts to use said material to challenge the terms of debate within the current concern with ‘essentialism’ in social theory. It is in complete agreement with constructivist-inspired observations by anthropologists that so-called ethnic groups in Mongolia exist partly as ideological ascriptions and fictitious claims arising from historical and contemporary political interests. Yet there is an intellectual tension, which sees the anthropologist engaged in the politically important task of the de-essentialisation of identities, on behalf of people who claim said identities as their essence. This paper neither accepts ethnic groups as somehow unproblematically real, and yet also does not render such ideas purely as a target for deconstruction. Rather, it provides an account of times when moralised understandings of ethnic difference in the medium of mythic-historical exemplars are mobilised and examines the discursive content of such claims in terms of their impact on social life. I argue that such a moral historicity is itself the key characteristic of ethnicity in Mongolia, one that has particularly clear implications for our concern with essentialism. It is clear that essentialism has (rightly) become a dirty word within anthropology. Yet this has had the effect of blocking a sustained consideration as to what it actually does in social life. This paper argues that certain forms of essentialism may contain affordances of divergent interpretation that amplify social and political possibilities, rather than violently close them down. The final provocation then, is that anthropologists and social theorists would do better to examine the precise nature and effects of particular instances of essentialism, ethnographic and theoretical, rather than instinctively demolish them in general.