A parallel randomized controlled trial compared metacognitive therapy (MCT) with prolonged exposure (PE) in 32 patients with PTSD of ≥3 months duration. Participants were assigned to; eight sessions of therapy (MCT or PE) or an 8-week wait period (WL). There was only one drop-out from each treatment. Both active treatments were effective, resulting in significantly lower symptoms of PTSD, anxiety and depression compared with the WL. At post-treatment MCT was superior to PE on self-report symptoms of PTSD and superior to WL on objective measures of hyper-arousal (heart-rate). Recovery rates and reliable improvement in both MCT and PE were high. MCT exerted effects more rapidly and within group effect sizes in MCT were much larger than those in PE at the end of treatment. Clinical gains remained evident at follow-up by which time the treated groups did not differ. In conclusion; both treatments were effective but MCT had a clear advantage. The clinical implications and suggestions for future research are discussed.