How can we ever know, unequivocally, that another person is aware? Notwithstanding deeper philosophical considerations about the nature of consciousness itself, the only reliable method we have for detecting awareness in others is by eliciting a predicted response to an external prompt or command. Logically therefore, our ability to detect awareness in others is determined, not by whether they are aware or not, but by their ability to communicate that fact through a recognized behavioural response. This problem exposes a central conundrum in the study of awareness in general, and in particular, how it relates to the vegetative state and other so-called ‘disorders of consciousness’. From this perspective, I will discuss various solutions to this conundrum using functional neuroimaging. In particular, I will contrast those circumstances in which imaging data can be used to infer awareness in the absence of a reliable behavioural response, with those circumstances in which it cannot.