Paranormal beliefs are held by a large proportion of the population, yet the existence of paranormal phenomena is not obvious. There have been four main approaches to explaining why some people believe and others do not, emphasising the attitudinal, demographic, personality, and cognitive correlates of paranormal belief, in turn. However, what has so far been absent from the literature is an adequate theoretical framework which integrates the findings of research under these four hypotheses into one coherent explanation. We set out to re-examine previously reported relationships in this area, and attempted to develop an integrated theoretical perspective. We measured paranormal beliefs, personality, intelligence, emotional intelligence, and schizotypy, using the Revised Paranormal Belief Scale (RPBS), the Big Five Inventory (BFI), a public domain IQ test, the Schutte Self-Report Emotional Intelligence Test (SSEIT), and the Short Scales for Measuring Schizotypy, respectively. We also obtained demographic information such as age and gender from respondents. Pearson’s correlation showed that Global Paranormal Belief was positively related to social marginality, Unusual Experiences, Emotional Intelligence, and Extraversion, and negatively related to intelligence. Multiple regression showed that these five factors accounted for 56% of the variance in Global Paranormal Belief scores. Mediation analysis indicated that social marginality and Unusual Experiences mediated the other three factors. On the basis of this, we propose a new, integrative account of paranormal belief, the fantasy-reality mismatch hypothesis. According to this hypothesis, paranormal believers experience their own fantasy world which does not always match reality.