This study was reviewed and approved by the University of Sheffield Ethics Committee (039900).  Dataset for the CBTS-ED ratings from expert raters (experienced therapists working in the field of eating disorders) and non-expert raters (less experienced in the field of eating disorders, e.g., trainee therapists).  Part A CBTS-ED items (10 items) should be rated every time. However, in the case of error in completion, up to two items can be missing and the scale can still be scored, using pro-rating to yield an overall item mean score. Part B items (16 items) are rated only when relevant to the session being observed. The overall score for Part B is the item mean score for the number of items identified as relevant and scored. Six mock CBT-ED sessions ("vignettes") were observed and rated...