Doctor of Philosophy (PhD) / The purpose of this thesis was to investigate relationships between treatment knowledge, beliefs and outcome in Panic Disorder and/or Agoraphobia (Panic-Ag). Research from the psychotherapy and medical literature indicates patients’ treatment knowledge and beliefs, specifically acceptance of the treatment rationale (ATR), expectancies of treatment outcome (ETO) and treatment self-efficacy (TSE), are associated with clinical outcomes for a range of disorders. However, methodological limitations surrounding measurement of these constructs have undermined conclusions and/or such relationships have not been investigated in the field of Panic-Ag. Relationships between treatment knowledge, beliefs and outcome in Panic-Ag were examined using a 2 phase procedure. Phase 1 involved developing measures of treatment knowledge, ATR, ETO and TSE using patient and clinician samples. The psychometric properties of these measures were found to be satisfactory. Phase 2 investigated associations between treatment knowledge, beliefs and outcome following cognitive behaviour therapy (CBT) among 41 Panic-Ag participants. Measures were administered at pretreatment and 6-months posttreatment. It was hypothesised that treatment knowledge, ATR, ETO and TSE would be related to outcome, with associations mediated by belief in catastrophic cognitions. Of 4 Panic- Ag outcome measures (panic attack frequency, panic sensation severity, frequency of catastrophic cognitions and agoraphobic avoidance), results indicated improved treatment knowledge was significantly associated with frequency of catastrophic cognitions and agoraphobic avoidance. Posttreatment TSE was significantly associated with panic attack sensation severity, frequency of catastrophic cognitions and agoraphobic avoidance. Contrary to the hypothesis, ATR was not related to outcome. Similar findings concerning TSE and ATR were obtained in an independent sample of 34 Panic-Ag participants. Exploratory analyses found that pretreatment beliefs including outcome expectancies were unrelated to outcome. Mediational analyses revealed relationships between TSE and outcome were partially mediated by belief in catastrophic cognitions while relationships between treatment knowledge and outcome were not. Results are discussed in light of previous research, methodological limitations, clinical implications and future research directions.