There is growing evidence that sleep disturbances may impede the utility of existing therapeutic interventions for people with posttraumatic stress disorder (PTSD). This retrospective medical record review examined the hypothesis that sleep disturbance affects the outcome of prolonged exposure (PE) therapy for PTSD. We identified 18 combat veterans with PTSD who had completed PE therapy. There were 6 subjects who had sleep‐disordered breathing, 5 of whom were documented by sleep polysomnography. All subjects in the sleep‐disordered group took part in a minimum of 10 sessions; the mean number of sessions was comparable between the sleep‐disordered group and the group without a sleep disorder. Posttreatment PTSD Checklist scores were significantly reduced in those without a sleep disorder (−28.25; 58.0% reduction, F(1, 11) = 59.04, p < .001), but were not reduced in those with sleep‐disordered breathing (−7.17; 13.5% reduction, d_IGPP = 2.25 [independent groups pretest‐posttest design]). These observations supported the hypothesis that the efficacy of PE therapy is affected by sleep quality. If these findings are replicated, treatment algorithms may need to incorporate the presence or absence of sleep disorders as a factor in treatment choice.