Abstract: The study of the effects of war exposure on the psychological health of combatants has so far been constrained by possible selection biases which limits the establishment of causality, the clear identification of dynamics, and the generalizability of findings. In this study, we make use of a population-level natural experiment enabled by the strict military conscription system in Turkey which uses a draft lottery to randomly allocate conscripts to bases across the country, including those south-eastern areas experiencing a long running civil conflict. We build on this setting with a representative field survey of 5024 adult males. Our results indicate that those exposed to high intensity armed conflict environments during their service are more likely to experience depressive symptoms even long after their discharge. Further detailing conflict exposure, we find war traumas to be the primary drivers of the effects we observe.