Previous research into paranormal belief has traditionally been based upon questionnaires and belief scales (e.g. Tobacyk, 2004). However, there are a number of limitations associated with these types of methods with regard to the study of beliefs and attitudes. As such, this study uses a discursive approach in the study of fact construction and paranormal belief. Previous discursive research into the paranormal has focused either on how individuals construct their accounts as factual (Wooffitt, 1992) or on how individuals warrant their belief in the paranormal (Lamont, 2007b; Lamont et al. 2009). This project uses discourse analysis to investigate how accounts of ghostly experiences are constructed. It aims to show how individuals justify the facticity of their account and also justify their belief that the events were paranormal in nature. Three common themes were identified from the data: 1) individuals constructing their accounts to show that they are reliable witnesses; 2) labelling the events as paranormal by means of group consensus; and 3) constructing events to emphasise clear observation. Individuals draw upon a variety of discursive strategies in order make their accounts sound factual and justify their belief that the events were paranormal. These strategies, in line with previous discursive literature, include the use of consensus (Edwards & Potter, 1992), extreme-case formulations (Pomerantz, 1986), three-part lists (Jefferson, 1990), stake inoculation and interest (Potter, 1996), and avowals of prior scepticism (Lamont, 2007b). The findings of the study are then discussed before future directions for the discursive study of paranormal belief are suggested.