Background: Eating disorders (EDs) affect an increasing proportion of young women in western countries. Psychometric questionnaires represent valuable tools to investigate various and critical areas directly involved in the pathogenesis of EDS and to support diagnosis and therapeutic decisions. Methods: 162 young women (16–35 years old) seeking diettherapy were recruited. We classified subjects in normal eating behaviour (NEB) (n=87), binge eating disorder (BED) (n=12) and bulimic EDNOS (Eating Disorders not Otherwise Specified) (n=63). The SCOFF, Eating Attitudes Test (EAT 26) and Three Factor Eating Questionnaire (TFEQ) were administered. Body mass index (BMI) was utilised to assess the nutritional status. An analysis of the reliability and validity (sensitivity and specificity) of the SCOFF, EAT 26 and TFEQ was performed. Results: Body mass index (BMI) of NEB, BED and bulimic EDNOS was 27.7, 35 and 31.1, respectively. BED showed the highest values at the dishinibition, hunger and food preoccupation scales but, conversely, they were the least restrained group. The SCOFF was significantly associated with the dishinibition (r=0.31), hunger (0.31), dieting (r=0.34) and food preoccupation scales (r=0.34). The reliability analysis showed that the SCOFF, EAT 26 and TFEQ had a Cronbach alpha of 0.47, 0.85 and 0.75, respectively. The ROC curves identified cut off points of 3, 10 and 25 as the best compromise between specificity and sensitivity for the SCOFF, EAT 26 and TFEQ, respectively. Conclusions: The SCOFF is a valuable tool for the screening of abnormal eating behaviours but the diagnosis should be always confirmed and supported by the administration of other questionnaires and structured interviews. We have also confirmed the high reliability of the EAT 26 and TFEQ, even though the utilisation of these questionnaires has generated some issues about their application in populations characterised by loss of control and overeating episodes.