Current conceptions of borderline personality disorder, like those of other forms of personality disorder, are largely derived from clinical concepts and observations. Over the last decade or so, however, empirical analyses of the structure of personality disorder and etiological research have begun to offer a new understanding of the structure and origins of personality disorder, including borderline pathology, creating the possibility of developing diagnostic constructs that more closely match empirical findings. The Diagnostic and Statistical Manual of Mental Disorders (4th ed.) (DSM-IV) concept of borderline personality disorder as a discrete diagnostic category is challenged by evidence of extensive overlap with other personality disorders. The implication is that DSM-IV concepts of personality disorder do not reflect natural distinctions in the way personality pathology is organized. This assertion is supported by statistical analyses of the descriptive features of the personality disorders that have consistently failed to identify factors that resemble DSM concepts (Austin & Deary, 2000; Ekselius, Lindstrom, von Knorring, Bodlund, & Kullgren, 1994). These findings raise questions about the validity of DSM-IV personality disorder diagnoses, including borderline.