This paper discusses the role of consent in decision making generally and its role in end of life decisions in particular. It outlines a conception of autonomy which explains and justifies the role of consent in decision making and criticises some misapplications of the idea of consent, particular the role of fictitious or "proxy" consents. Where the inevitable outcome of a decision must be that a human individual will die and where that individual is a person who can consent, then that decision is ethical if and only if the individual consents. In very rare and extreme cases such a decision will be ethical in the absence of consent where it would be massively cruel not to end life in order to prevent suffering which is in no other way preventable. Where, however, the human individual is not a person, as is the case with abortion, the death of infants like Mary (one of the conjoined twins in a case discussed in the paper), or in the very rare and extreme cases of those who have ceased to be persons like Tony Bland, such decisions are governed by the ethics of ending the lives of non-persons.