Although a number of effective psychotherapeutic treatments have been developed for borderline personality disorder (BPD), little is known about the mechanisms of change explaining the effects of these treatments. There is increasing evidence that impairments in mentalizing or reflective functioning—the capacity to reflect on the internal mental states of the self and others—are a central feature of BPD. To date, no study has directly investigated the core assumption of the mentalization-based approach to BPD, that changes in this capacity are associated with treatment outcome in BPD patients. This study is the first to directly investigate this assumption in a sample of 175 patients with BPD who received long-term hospitalization-based psychodynamic treatment. Using a parallel process growth modeling approach, this study investigated whether (a) treatment was related to changes in mentalizing capacity as measured with the Reflective Functioning Questionnaire; (b) these changes could be explained by pretreatment levels of mentalizing and/or symptomatic distress; and (c) changes in mentalizing capacity over time were associated with symptomatic improvement. Mentalizing and symptomatic distress were assessed at admission, 12 and 24 weeks into treatment, and at discharge. Results showed that treatment was associated with significant decreases in mentalizing impairments (i.e., uncertainty about mental states) and symptomatic distress. Pretreatment levels of mentalizing and symptomatic distress did not predict these changes. However, improvements in mentalizing were strongly associated with the rate of decrease in symptomatic distress over time (r = .89). These findings suggest that increases in mentalizing may indeed in part explain therapeutic change in the treatment of BPD, but more research is needed to further substantiate these conclusions.