The current thesis sought to investigate whether or not our ability to keep track of reality known as ‘reality monitoring’ may depend on the qualitative characteristics of underlying mnemonic representations, such as the precision with which we remember details of past experiences (Johnson et al., 1993). To test this notion, experiments reported in the current thesis used behavioural measurements, brain stimulation, cognitive manipulations, and measures of individual differences. Data from these experiments, by and large, suggests that reality monitoring decisions may be associated with the precision of recollection, where greater reality monitoring performance tend to track greater precision of recollection. If, however, the precision of recollection is reduced, reality monitoring decisions may instead draw on a false sense of familiarity, as opposed to recollection, inducing a tendency to misattribute imagined experiences as those perceived from the outside world. Even people with, on average, greater precision of recollection tended to exhibit both behavioural and personality traits often observed in people with a tendency to lose touch with reality or ‘psychosis’, although this association did not survive stringent statistical corrections for false positive rates, suggesting that better mnemonic qualities may not always help us to keep track of reality. On balance, these findings suggest that multiple relationships may exist between reality monitoring decisions and the underlying mnemonic qualities. Such multiple relationships might be better described by a combination of previously separate theoretical accounts, rather than any single theoretical account. In light of these theoretical implications, the current thesis proposes a combined model that aims to account for the multiple relationships between qualities of mnemonic representation, reality monitoring decisions, and traits of psychosis.