Whether or not the use of maladaptive defense style is a trait, as opposed to a state dependent phenomenon, in panic disorder (PD) is a topic still very much up for debate. The aim of the study was to verify whether PD patients, both before and after treatment, used different defense style than the control group. Sixty-one PD patients (recruited from an original sample of 90 patients) and 64 healthy controls were evaluated against the Structured Clinical Interview for DSM-IV disorders, the Symptoms Check List-90, the Hamilton Rating Scales for Anxiety and for Depression and finally the Defense Style Questionnaire-40 (DSQ). The patients were treated with paroxetine or citalopram and were evaluated monthly for one year to assess the remission. The DSQ was re-administered to the patients at the end of the study. Before treatment, PD patients used more neurotic and immature forms of defense than controls. After treatment, those in remission used the same defense styles as the control group, whereas non-remitters still used more immature defenses. However, all the aforementioned difference disappeared, after excluding the effect of symptom severity. Our data supports the hypothesis that the use of maladaptive defenses might be the consequence of PD: when subjects fall ill, their capacity to use mature adaptive defenses may diminish, but when they recover their defensive style returns to a greater maturity. The present results are however limited by the dropout rate (one third of patients did not complete the study) and the use of just one questionnaire to evaluate the complexity of defense styles.