This paper addresses the newly named phenomenon of ‘rescue supervision’. The phenomenon itself is unfortunately not new. Not every candidate is blessed with adequate doctoral supervision. This paper looks at some of the reasons why supervision may not be adequate, only one of which is that the supervisor is incompetent. Whatever the reasons for the inadequacy of the supervision, it is desirable that the situation be faced and action taken. Sometimes it is enough to change the supervision team, but in other cases damage may have been done to the progress of the research and to the candidate, and in such cases change is not enough. In this latter case the situation must be rescued by a ‘rescue supervisor’. Such a supervisor has to bring together two aspects: the ordinary competencies of supervising a project, and the extra-ordinary requirement to take on a damaged candidate and project and deal with the consequences. A rescue supervisor is appointed to a failing PhD. The appointment is different from the normal process of changing supervision owing to staff turnover, etc. It is a process that is forced upon the institution, or the candidate who has become the victim of circumstances. I have been appointed as a rescue supervisor four times, and this paper is based on my experiences of that process compared to my experiences of normal doctoral supervision and examination.