Delphi research is an increasingly popular method within medical, social and psychological research. It tends to be employed where established theory or knowledge are lacking but where 'experts' are thought to hold relevant information. It consists of developing and administering sequential questionnaires that seek to move towards a position of relative consensus. Although the original authors of the technique established a specific method, the literature reveals modifications in the way this is applied. Variations include (i) restricting the ability of experts to respond to the original question, (ii) changing or varying the expert groups used and (iii) the point at which the research ends. This paper provides an overview of the technique and explains these variations and their implications before highlighting possible ways forward.