Some psychometric properties of the Bulimic Investigatory Test, Edinburgh (BITE; Henderson and Freeman, 1987) were analysed in a non-clinical sample. The questionnaire was administered to 1122 adolescents, 669 girls and 453 boys from different schools, aged 12 to 21, from Malaga (Spain). The BITE consists of two subscales: the Symptom Scale and the Severity Scale. Both scales showed a unidimensional structure and a high internal consistency. From a ROC curve, cut-off points were established in the BITE scores related to DSM-IV bulimia diagnostic (American Psychiatric Association, 1994). With these cut-off points, a high specificity was found, but a lower sensitivity than that reported in clinical samples. However, scores in the bulimic group were higher than in any other eating disorder group or in those with no disorder. Thus, BITE shows a certain degree of validity in assessing specific symptoms of bulimia. As a diagnostic instrument, however, it may give a high rate of false negatives in samples from general populations.