In this paper I argue for the existence of non-symmetric relations. Armstrong and Dorr have argued that (1) non-symmetric relations are trouble, and (2) that we don’t need non-symmetric relations because we can make do with symmetric relations when it comes to accounting for the order we find in nature. But, I argue, Armstrong and Dorr are wrong about both (1) and (2). Their efforts to do without non-symmetric relations presuppose them, whilst their arguments that non-symmetric relations are subject to a tu quoque response. This paper contributes a further plank to the ostrich realism about relations that I advocate in Macbride (2013).