Although autonomy is clearly still the paradigm in bioethics, there is increasing concern over its value and feasibility. In agreeing with those concerns, I argue that autonomy is not just a status, but a skill, one that must be developed and maintained. I also argue that nearly all health-care interactions do anything but promote such decisional skills, since they rely upon assent, rather than upon genuinely autonomous consent. Thus, throughout most of their medical lives, patients are socialised to be heteronomous, rather than autonomous. Yet, at the worst possible time--critical care decision-making--when life and death consequences are attached to the choices, the paradigm shifts and real consent is sought, even demanded, thereby making an often traumatic situation even harder. I go on, though, to also reject paternalistic models of beneficence as an alternative. Rather, I conclude that the problem is so fundamental in healthcare that a genuine solution would require a radical restructuring. I recommend steps that can be taken in the interim to improve the situation and to move toward such a restructuring.