forming trusting therapeutic relationships, often against a background of severe interpersonal trauma-should not be underestimated. As argued by Fonagy and colleagues, avoiding making a diagnosis on this basis could reinforce negative stereotypes of BPD. Nonetheless, this does place a responsibility on the mental health professionals firstly to be sure of the validity and utility of making such a diagnosis, and secondly to do as much as we can to combat the stigma associated with personality disorder.Fonagy and colleagues highlight two difficulties with our current systems for diagnosing personality disorder that raise questions regarding its validity in both adulthood and adolescence. Firstly, Fonagy and colleagues recommend integrating dimensional factors alongside categorical criteria when making a diagnosis of BPD in adolescents. This reflects a growing recognition of the questionable diagnostic validity of personality disorders as distinct categorical entities. In adults, supposedly distinct personality disorders are often highly comorbid, and factor analyses of personality disorder traits yield inconsistent results, with factor structures that do not resemble the distinct disorders outlined in 9,11,13,15]. The factor analytic structure of personality disorder traits in adolescents has been less well studied, with mixed findings. Using exploratory factor analysis of clinician ratings of DSM-IV personality disorder traits in a sample of 296 adolescent inpatients, Durrett and Westen [3] identified a factor structure resembling the ten DSM-IV personality disorders. Conversely, when a Q-factor analysis-which identifies clusters of similar people rather than clusters of co-occurring items-was applied to the same dataset, very different findings emerged [18]. Borderline personality disorder correlated with two empirically distinct and non-overlapping factors comprising emotional dysregulation and histrionic traits. These findings suggest that the distinctions between personality