Sharon Street (2006) has argued against ethical realism and for ethical constructivism with, what she calls, the "Darwinian Dilemma." This dilemma pushes the ethical realist to grab either of two unacceptable horns. The horns are that there is a relation between evolutionary influences and our evaluative attitudes and independent evaluative truths, or that there is no such relation. The first horn runs into problems with evolutionary psychology while the second horn seems to undermine ethical realism. I consider recent objections to this argument by Copp, Wielenberg, Rabinowitz, and Skarsaune and show that they are either ineffective against the dilemma or that there are plausible reconstructions of the dilemma which are immune to the objections. With the reinforced dilemma on the table I survey some implications for it from evolutionary psychology (or rather, the objections to it!) and then argue that the dilemma poses a serious problem to naturalistic theories of ethical realism. The upshot of this argument is that if we have excellent reasons for thinking that ethical realism is true, then we will need to appeal to or construct theories of ethical realism which are non-naturalistic.