Borderline personality disorder (BPD) is defined by the presence of at least five of nine symptoms in Section II of the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5). In the DSM-5 Section III Alternative Model of Personality Disorders (AMPD), BPD is defined by deficits in self and/or interpersonal functioning (Criterion A), elevated negative affectivity, and elevated antagonism and/or disinhibition (Criterion B). However, it is unclear if these definitions describe the same people and if the AMPD criteria explain unique variability in treatment outcomes in this population. In a treatment-seeking sample of adult participants diagnosed with BPD according to Section II criteria (n = 65; Mage = 27.60, 70.8% female, 76.9% white), we found a majority (66.2%) would have also received the diagnosis based on AMPD criteria. Those meeting AMPD criteria reported more severe Section II BPD symptoms than those who did not, ps < .02, ds > .60, and the presence or severity of Section II fears of abandonment and inappropriate anger uniquely predicted AMPD BPD diagnoses, ps < .03, ORs > 2.31. Changes in AMPD dimensions explained 34% of the variability in change in work/social adjustment (p = .13) and quality of life (p = .22), respectively, over and above changes in Section II symptoms during a novel cognitive-behavioral treatment for BPD. These results suggest AMPD criteria capture a more severe subset of BPD than Section II criteria and may be important predictors of treatment outcomes. We discuss the potential trade-offs of this shift in diagnosis.