Greene uses evidence from psychology and neuroscience to argue that manual mode (slow, deliberative thinking) is conducive to utilitarian judgments. He further argues that these data, in conjunction with philosophical premises, lend normative support to utilitarianism. After defending Greene’s philosophical premises against critics, I contend that the current state of the evidence suggests that manual mode does not drive utilitarian responses to moral dilemmas involving self-sacrifice. I performed an experiment which replicated the positive association between cognitive reflection test (CRT) scores (which measure reliance on manual mode) and utilitarian responses to dilemmas that involved sacrificing the interests of others. However, I did not find a positive association between CRT scores and self-sacrificial utilitarian responses. The lack of a connection between manual mode and self-sacrifice presents a problem for Greene’s argument that manual mode drives utilitarianism in general. Prima facie, my results indicate that reflection only drives other-sacrificial utilitarian judgments, not self-sacrificial ones. Greene is left without a basis to say (as he does) that cognitive science lends support to the normative conclusion that we ought to engage in utilitarian self-sacrifice by, for example, giving more to charity. I conclude by discussing other implications of my data for Greene’s argument, and outlining directions for future research.