Background: Our goal was to better understand distinct PTSD symptom presentations in Iraq and Afghanistan Veterans (N=227) and to determine whether those who killed in war were at risk for being in the most symptomatic class. Methods: We used latent class analysis of responses to the PTSD checklist and logistic regression of most symptomatic class. Results: We found that a four-class solution best fit the data, with the following profiles emerging: High Symptom (34% of participants), Intermediate Symptom (41%), Intermediate Symptom with Low Emotional Numbing (10%), and Low Symptom (15%). The largest group of individuals who reported killing (45%) was in the High Symptom class, and those who killed had twice the odds of being in the most symptomatic PTSD class, compared to those who did not kill. Those who endorsed killing a non-combatant (OR=4.56, 95% CI [1.77, 11.7], p<0.01) or killing in the context of anger or revenge (OR=4.63, 95% CI=[1.89, 11.4], p<0.001) were more likely to belong to the most symptomatic PTSD class, compared to those who did not kill. Limitations: The study was retrospective and cross-sectional. The results may not generalize to veterans of other wars. Conclusions: Killing in war may be an important indicator of risk for developing frequent and severe PTSD symptoms. This has implications for the mental healthcare of veterans, providing evidence that a comprehensive evaluation of returning veterans should include an assessment of killing experiences and reactions to killing. © 2012 Elsevier B.V.