Prinz claims that empirical work on emotions and moral judgement can help us resolve longstanding metaethical disputes in favour of simple sentimentalism. I argue that the empirical evidence he marshals does not have the metaethical implications he claims: the studies purporting to show that having an emotion is sufficient for making a moral judgement are tendentiously described. We are entitled to ascribe competence with moral concepts to experimental subjects only if we suppose that they would withdraw their moral judgement on learning that they were fully explained by hypnotically induced disgust. Genuine moral judgements must be reason-responsive. To capture the reason-responsiveness of moral judgement, we must turn to either neo-sentimentalism or to a non-sentimentalist metaethics, either of which is fully compatible with the empirical evidence Prinz cites.