Cognitive failures are minor errors in thinking reported by clinical and non-clinical individuals during everyday life. It is not yet clear how subjectively-reported cognitive failures relate to objective neuropsychological ability. We aimed to consolidate the definition of cognitive failures, outline evidence for the relationship with objective cognition, and develop a unified model of factors that increase cognitive failures. We conducted a systematic review of cognitive failures, identifying 45 articles according to the PRISMA statement. Failures were defined as reflecting proneness to errors in 'real world' planned thought and action. Vulnerability to failures was not consistently associated with objective cognitive performance. A range of stable and variable factors were linked to increased risk of cognitive failures. We conclude that cognitive failures measure real world cognitive capacity rather than pure 'unchallenged' ability. Momentary state may interact with predisposing trait factors to increase the likelihood of failures occurring. Inclusion of self-reported cognitive failures in objective cognitive research will increase the translational relevance of ability into more ecologically valid aspects of real world functioning.