Over 30% of combat injuries involve peripheral nerve injury compared to only 3% in civilian trauma. In fact, nerve dysfunction is the second leading cause of long-term disability in injured service members and is present in 37% of upper limb injuries with disability. Identification and assessment of non-penetrating nerve injury in trauma patients could improve outcome and aid in therapeutic monitoring. We report the use of Raman spectroscopy as a noninvasive, non-destructive method for detection of nerve degeneration in intact nerves due to non-penetrating trauma. Nerve trauma was induced via compression and ischemia/reperfusion injury using a combat relevant swine tourniquet model (>3 hours ischemia). Control animals did not undergo compression/ischemia. Seven days post-operatively, sciatic and femoral nerves were harvested and fixed in formalin. Raman spectra of intact, peripheral nerves were collected using a fiber-optic probe with 3 mm diameter spot size and 785 nm excitation. Data was preprocessed, including fluorescence background subtraction, and Raman spectroscopic metrics were determined using custom peak fitting MATLAB® scripts. The abilities of bivariate and multivariate analysis methods to predict tissue state based on Raman spectroscopic metrics are compared. Injured nerves exhibited changes in Raman metrics indicative of 45% decreased myelin content and structural damage (p<<0.01). Axonal and myelin degeneration, cell death and digestion, and inflammation of nerve tissue samples were confirmed via histology. This study demonstrates the non-invasive ability of Raman spectroscopy to detect nerve degeneration associated with non-penetrating injury, relevant to neurapraxic and axonotmetic injuries; future experiments will further explore the clinical utility of Raman spectroscopy to recognize neural injury.