Background: In this prospective study, Personality Disorders (PersD) were evaluated in patients with Panic Disorder (PD), before and after one year of pharmacotherapy to verify whether personality characteristics changed after treatment. Method: Sixty PD patients and 60 sex and age-matched normal controls participated in the study. All subjects were evaluated with the SCID-IV, the Structured Interview for DSM-IV Personality Disorders (SIDP), the SCL-90, the Ham-A and the Ham-D. Patients were treated with paroxetine or citalopram and were evaluated monthly for one year to assess the remission of symptoms. The SIDP was re-administered to the patients at the end of the study. Results: Before treatment, PD patients showed a higher prevalence (60%) of PersD than normal subjects (8%). After treatment, PersD rate decreased (43%) due to the reduction of the rate of paranoid, avoidant and dependent PersD. When the effect of the treatment on personality traits was evaluated, we found that avoidant traits decreased only in remitted patients, paranoid traits decreased both in remitted and in non-remitted patients, and dependent traits decreased only in patients with major depression comorbidity. Limitations: The small sample size and the short length of the follow-up period of our study suggest caution in the generalization of our results. Conclusions: In our PD patients, an improvement of symptoms was associated with a reduction of paranoid, avoidant and dependent traits, with a normalization of paranoid traits and a persistence of avoidant and dependent characteristics. Therefore, our data suggest that in PD patients not only paranoid traits but also avoidant and dependent traits show, at least in part, a state phenomenon. (c) 2005 Elsevier B.V. All rights reserved.