The current study investigated changes in service members' cognitions over the course of Cognitive Processing Therapy (CPT) for posttraumatic stress disorder (PTSD). Sixty-three active duty service members with PTSD were drawn from 2 randomized controlled trials of CPT-Cognitive Only (CPT-C). Participants wrote an impact statement about the meaning of their index trauma at the beginning and again at the end of therapy. Clauses from each impact statement were qualitatively coded into three categories for analysis: assimilation, accommodation, and overaccommodation. The PTSD Checklist, Posttraumatic Symptom Scale-Interview Version, and the Beck Depression Inventory-II were administered at baseline and posttreatment. Repeated measures analyses documented a significant decrease in the percentage of assimilated or overaccommodated statements and an increase in the percentage of accommodated statements from the beginning to the end of treatment. Changes in accommodated statements over the course of treatment were negatively associated with PTSD and depression symptom severity, while statements indicative of overaccommodation were positively associated with both PTSD and depression symptom severity. Treatment responders had fewer overaccommodated and more accommodated statements. Findings suggest that CPT-C changes cognitions over the course of treatment. Methodological limitations and the lack of association between assimilation and PTSD symptom severity are further discussed.