Objective: Individuals meeting all criteria for anorexia nervosa (AN) except that weight falls within or above the normal range despite significant weight loss are categorized as having atypical AN (AAN). Existing research has provided mixed evidence concerning the diagnostic demarcation of AN and AAN. The aim of the present study was to identify research priorities for furthering the understanding of AN and AAN as diagnostic entities. Method: Employing the Delphi methodology, experts in the field were invited to suggest research questions that need to be explored in the demarcation of AN from AAN. This yielded 24 research areas, that were presented in subsequent rounds where panelists were asked to prioritize areas of primary interest. Results: Fifty-three panelists completed all three Delphi rounds. Consensus was only reached on three items considered to be of primary interest: medical, neurobiological, and neurological factors; epidemiology and natural course; and treatment response in AAN compared to AN. In contrast, questions of premorbid weight and determining the need for and nature of a body mass index cutoff differentiating between AAN and AN were seen as being of low priority. Discussion: These findings reveal a relatively low degree of consensus on the demarcation of AN from AAN in the field of eating disorders. A reason could be that the definition and use of the AAN category vary in research and clinical practice. In order to achieve further diagnostic clarity, research on the demarcation of AAN and AN should focus on the identified prioritized research areas.