The current study aimed to establish whether anxiety predicts subsequent anorexia nervosa onset and maintenance. A systematic review of longitudinal studies assessing the association between stable anxiety exposures (e.g. trait anxiety/anxiety disorder pathology) and anorexia nervosa development or maintenance was undertaken. Eight studies met inclusion criteria. Seven probed the association between anxiety and anorexia nervosa onset, and one assessed the association between anxiety and anorexia nervosa maintenance. Individuals with anorexia nervosa were more likely to report childhood anxiety compared to healthy individuals, but whether childhood anxiety explains unique variance in anorexia nervosa development is unclear. Current evidence does not support longitudinal associations between specific anxiety disorders (independently of other anxiety disorders) and subsequent anorexia nervosa onset, however anxiety disorder diagnosis in general may predict increased anorexia nervosa risk. The single study probing the association between anxiety and anorexia nervosa maintenance did not find evidence supporting a relationship. The quality of individual studies was fair to high, however the body of evidence was of low quality. Further research that minimises bias, allowing for strong conclusions concerning longitudinal associations between anxiety and subsequent anorexia nervosa outcomes, is required to inform anorexia nervosa aetiology. This in turn may promote improved prevention and treatment.