Residential posttraumatic stress disorder (PTSD) research in military samples generally shows that in aggregate, PTSD symptoms significantly improve over the course of treatment but can remain at elevated levels following treatment. Identifying individuals who respond to residential treatment versus those who do not, including those who worsen, is critical given the extensive resources required for such programs. This study examined predictors of treatment response among 282 male service members who received treatment in a U.S. Department of Defense residential PTSD program. Using established criteria, service members were classified as improved, indeterminate (referent), or worsened in terms of self‐reported PTSD symptoms. Multinomial logistic regression results showed that for PTSD symptoms, higher levels of pretreatment PTSD symptom severity were associated with significantly lower odds of being in the improved group, adjusted odds ratio (aOR) = 0.955, p = .018. In addition, service members who completed treatment were significantly more likely to be in the improved group, aOR = 2.488, p = .048. Longer average pretreatment nightly sleep duration, aOR = 1.157, p = .035, and more severe pretreatment depressive symptoms, aOR = 1.109, p = .014, were associated with significantly higher odds of being in the improved group. These findings reveal clinical characteristics better suited for residential PTSD treatment and highlight implications for comorbid conditions.