This article examines the ability of the panic attack-posttraumatic stress disorder (PTSD) model to predict how panic attacks are generated and how panic attacks worsen PTSD. The article does so by determining the validity of the panic attack-PTSD model in respect to one type of panic attack among traumatized Cambodian refugees: orthostatic panic (OP) attacks (i.e. panic attacks generated by moving from lying or sitting to standing). Among Cambodian refugees attending a psychiatric clinic, the authors conducted two studies to explore the validity of the panic attack-PTSD model as applied to OP patients (i.e. patients with at least one episode of OP in the previous month). In Study 1, the panic attack-PTSD model accurately indicated how OP is seemingly generated: among OP patients (N = 58), orthostasis-associated flashbacks and catastrophic cognitions predicted OP severity beyond a measure of anxious-depressive distress (Symptom Checklist-90-R subscales), and OP severity significantly mediated the effect of anxious-depressive distress on Clinician-Administered PTSD Scale severity. In Study 2, as predicted by the panic attack-PTSD model, OP had a mediational role in respect to the effect of treatment on PTSD severity: among Cambodian refugees with PTSD and comorbid OP who participated in a cognitive behavioural therapy study (N = 56), improvement in PTSD severity was partially mediated by improvement in OP severity.