Posttraumatic stress disorder (PTSD) is a trauma and stressor-related disorder that results in complex somatic, cognitive, affective and behavioural effects, after exposure to traumatic event(s). Conventional imaging (T1 and T2 weighted magnetic resonance imaging) has little to offer in the way of diagnosis of mental health conditions such as PTSD and there is currently no objective diagnostic test available. Magnetic resonance spectroscopy (MRS) allows for non-invasive measurement of metabolites and neurochemicals in the brain using a conventional MRI scanner and offers the potential to predict, diagnose and monitor PTSD. This systematic review summarises the results of 24 MRS studies, performed between 1998 and 2017, to measure neurochemical differences, occurring as a consequence of PTSD. The most consistent finding in subjects with PTSD is lower N-acetylaspartate levels in the hippocampus and anterior cingulate cortex, with and without atrophic change. More recent studies, using more advanced techniques and modern hardware, have shown evidence of glutamatergic dysfunction and differences in gamma-aminobutyric acid levels in the brain of patients with PTSD. Conflicting results have been reported in choline-containing metabolites and there is emerging evidence of glutathione being affected. Myo-inositol and creatine are unchanged in the majority of studies.