<p>This exploratory study aimed to examine the relationship between therapist-reported mentalizing capacity and self-reported personality pathology, specifically identifying distinct clusters based on self-reported dimensional personality pathology and evaluating differences in therapist-reported mentalizing across these clusters. Ninety treatment-seeking clients (67 women; M_age = 30.79) and their 22 therapists participated. Clients completed the Personality Inventory for DSM-5 and the Self and Interpersonal Functioning Scale. Therapists rated clients’ mentalizing capacities using the Mentalization Imbalances Scale and the Modes of Mentalization Scale. Hierarchical cluster analysis was used to identify distinct personality pathology profiles, and multilevel mixed-effects models were employed to examine differences in mentalizing capacities between clusters. A two-cluster solution emerged, with clients categorized into low-pathology (<i>N</i> = 45) and high-pathology (<i>N</i> = 45) groups. The high-pathology cluster demonstrated significantly higher levels across all personality pathology domains. Mixed-effects models revealed that clients in the high-pathology cluster showed significantly lower Good Mentalization and higher impairments in several mentalizing domains, including Automatic, Self, Excessive Certainty, Affective, and Teleological Thought, with small-to-medium effect sizes (η² = 0.04–0.06). No significant between-cluster differences were found for Cognitive, External, Concrete Thinking, and Others domains. Thus, findings partially support the hypothesis that individuals with more severe personality pathology demonstrate pronounced mentalizing impairments. The pattern of deficits, with preserved cognitive/external mentalizing alongside impaired self-related, affective, and automatic processing, supports the Alternative Model for Personality Disorders, as these specific mentalizing impairments largely correspond with the self-related dysfunctions described in Criterion A.</p>