• A study was designed to determine whether the Diagnostic Interview for Borderlines (DIB) might be scored from the Schedule for Affective Disorders and Schizophrenia (SADS), and also whether DIB scores predicted the clinical diagnosis of DSM-III borderline personality disorder. One pair of clinicians interviewed patients with the DIB, and another pair interviewed the same patients with a slightly modified version of the SADS. Both interviews diagnosed virtually the same patients as borderline according to the criteria of Gunderson and Singer. The sensitivity of DIB scores in predicting a DSM-III diagnosis of borderline was 70%, while the specificity was 90%; the intraclass correlation coefficient was .75. Although there is a substantial concordance, the disparity between the DSM-III and DIB systems of diagnosing borderline patients is sufficiently great to preclude the generalization of findings from studies employing one set of criteria to those employing the other.