There are two versions of the Cognitive Processing Therapy (CPT) protocol: one that includes a trauma narrative and one that does not. Despite both versions being used in clinical practice, to date, there has not been a comprehensive comparison of posttraumatic stress disorder (PTSD) treatment outcomes associated with these two protocols. The current study is a meta-analytic review of 29 randomized controlled trials of CPT+A (k = 13) and CPT-C (k = 16) in which we investigated whether there are differences in treatment outcome effect sizes and treatment dropout between the two protocol versions. Sample type (military/veteran versus non-military/non-veteran) as a moderator was examined, given that less robust PTSD treatment outcomes have been typically observed in military and veteran samples. Meta-regression analyses revealed that there were no significant differences between the two CPT protocol versions in pre-post PTSD treatment effect sizes, t(26) = -0.743, p = .463. As expected, studies that enrolled a military or veteran sample reported an average smaller symptom reduction (g = 0.95) than studies that enrolled a non-military or non-veteran sample (g = 1.41), t(26) = 2.48, p = .019. There were also no significant differences between the two CPT protocols for treatment dropout, t(26) = 1.69, p = .104, or between studies enrolling military/veteran and non-military/non-veteran samples t(26) = -1.10, p = .282. The findings support the use of either CPT protocol in practice, but also demonstrate that veterans and service members may be less responsive to either CPT protocol than other trauma samples.