Neil Manson and Onora O’Neill have recently defended an original theory of informed consent in their book Rethinking Informed Consent in Bioethics (2007). The development of their ‘waiver’ model is premised on the failings of the theory of informed consent as disclosure, which is rejected on two counts: firstly, the disclosure model’s implicit reliance upon a ‘conduit-container’ model of communication means that the regulatory requirements of informed consent can rarely be achieved; secondly, the model’s purported ethical justification via a principle of respect for patient autonomy is presented as being vacuous. Despite having laudable motivations for rethinking informed consent, I argue that their theory of informed consent as waiver can be criticised on similar grounds. In order to support this thesis I object that Manson and O’Neill’s developed theory of agential communication is too intricate to easily meet the demands of informed consent as waiver. Secondly, I show that the model appears to be implicitly reliant upon a principle of respect for patient autonomy. Hence, despite improving upon the doctrine of informed consent, the waiver model needs further elucidation in order to avoid the problems mounted against the disclosure model.