Impaired cognitive flexibility has been suggested as a risk factor for the development of anorexia nervosa (AN). The current study aimed to 1) investigate cognitive flexibility in people at various levels of risk of AN; and 2) compare people with a history of AN to people at different levels of risk of AN in cognitive flexibility. The sample comprised of 262 community participants (79% female) and 36 participants with a lifetime diagnosis of AN (97.2% female) aged between 18 and 64 years old. Participants completed self-report (the Depression Anxiety Stress Scale short-form version, the Eating Disorders Examination-Questionnaire, the Neuroticism Scale, and the Cognitive Flexibility Inventory) and neurocognitive (the Trail Making Test and the Wisconsin Card Sorting Test) assessments online to evaluate eating disorder symptoms, depression, neuroticism, and cognitive flexibility. Using a cluster analysis, participants were allocated into low-, medium-, and high-risk of AN groups (n = 88, 128, 46, and 36 respectively). Although high-risk participants self-reported significantly poorer cognitive flexibility than the other risk groups, performance on the neurocognitive tasks was similar across groups. Further, participants with lifetime AN reported significantly poorer cognitive flexibility than the low-risk group. People at high-risk of AN may perceive themselves to have poorer cognitive flexibility compared to those at a lower risk of AN. These results have implications for early identification of people at high-risk of AN.