ABSTRACT Philosophical theorizing on democracy typically assumes away the existence of nonhuman animals – even though animals are systematically affected by democratic decisions in morally relevant ways. This paper inspects under which circumstances, if at all, this omission can be justified. First, I distinguish between two possible explanations of this neglect: (i) that animals are neglected by democratic theorists due to speciesist biases, according to a debunking account, or (ii) that animals are deliberately idealized away, on a reconstructive approach. Largely, this paper is devoted to assessing when assuming away animals can function as a theoretically productive or practically helpful idealization when theorizing about democracy. In particular, I identify three different aims that may be served by idealizing animals away in democratic theory: (i) to understand what makes democracy valuable, legitimate, or authoritative, (ii) to assess how the pitfalls of really existing democracies might be fixed and (iii) to isolate what robust agents owe each other in democratic societies. I then argue that, whereas the third goal might be usefully advanced by idealizing animals away, the first two goals can only be suitably tackled if we take into account how democracies affect animals’ interests.