Background: Borderline personality disorder (BPD) is a psychiatric diagnosis characterized by emotional and behavioural instability, and impaired ability to maintain relationships. Previous research has demonstrated an association between BPD and insecure attachment style. It has been argued that BPD is a disorder of attachment, with insecure attachment being associated with inadequate capacity to represent mental states, or to “mentalize”. There is evidence that people with BPD are impaired in their capacity to mentalize in the context of attachment relationships. The term “mentalization” encompasses a broad range of processes including metacognition. There is a theoretical basis for metacognitive deficits in BPD. However, there is a lack of empirical evidence regarding the role of metacognition in BPD and its relationship to adult attachment style. Method: Participants with BPD were recruited from Community Mental Health Teams, Clinical Psychology and a Dialectal Behaviour Therapy (DBT) service within NHS Highland. A comparison group of participants without BPD were recruited from the Clinical Psychology service, having been referred for symptoms of depression. Both groups were administered the Relationship Scales Questionnaire (RSQ)(a self-report measure of attachment); and a short version of the Metacognitions Questionnaire (MCQ-30). Severity of clinical symptoms and current mood was assessed using the Clinical Outcomes in Routine Evaluation (CORE) and the Hospital Anxiety and Depression Scale (HADS). Results: Participants with BPD scored significantly higher than those without BPD on the attachment-anxiety and attachment-avoidance dimensions of the RSQ. The BPD group also endorsed MCQ-30 items more than the comparison group. There was a significant difference between the groups on the MCQ-30 total score and four of the five subscale scores. There were significant positive correlations between attachment dimension scores and metacognition subscales. The strongest associations were between attachmentanxiety and “uncontrollability and danger” and “need to control thoughts” subscales of the MCQ-30. Only metacognition was predictive of current mood and distress levels. Conclusions: The results of this study show that people with BPD report high attachment-avoidance and attachment-anxiety in their relationships, relative to a non-BPD, depressed comparison group. These findings are consistent with the existing literature regarding the profile of attachment in BPD. This study also found that people with BPD also have more maladaptive metacognitions than people with symptoms of depression. An association between self-reported adult attachment style and maladaptive metacognitiion was demonstrated in the present study. Maladaptive metacognitive strategies and beliefs potentially contribute to maintenance of depressed and anxious mood, as well as broader symptoms of distress.