Pharmacological treatment with prazosin and psychological treatment with imagery rehearsal therapy (IRT) are the two main treatments of posttraumatic nightmares. The American Academy of Sleep Medicine task force recently listed IRT as the recommended treatment for trauma-related nightmares and changed the recommendation of prazosin to 'may be used'. This new recommendation was based on a single prazosin trial and not on a meta-analytic review of all available trials. The current meta-analysis aims to fill this gap in the literature. Eight studies on IRT and seven studies on prazosin (N = 1.078) were analyzed based on the random effects model. Relative to control groups, prazosin had a moderate to large effect on nightmare frequency (g = 0.61), posttraumatic stress symptoms (g = 0.81), and sleep quality (g = 0.85). IRT showed small to moderate effects on nightmare frequency (g = 0.51), posttraumatic symptoms (g = 0.31), and sleep quality (g = 0.51). No significant differences in effect were observed between prazosin and IRT on any of these outcomes (all p's > 0.10). It is concluded that downgrading the recommendation of prazosin may be a premature decision and that the aggregated results in this meta-analysis clearly show efficacy of both treatments.