Traumatic Brain Injury (TBI) is the most common injury in recent military conflicts, with nearly 500,000 service members sustaining a TBI since 2000. Mild TBI (mTBI), or concussion, is by far the most common type of TBI and has been associated with long-term cognitive complaints and functional impairment. While clinical assessment of mTBI (i.e., MRI and performance-based cognitive testing) occasionally captures subtle abnormalities in the acute period following mTBI, these measurements lack the sensitivity to assess the time course of cognitive recovery from mTBI. The current study assessed cognitive changes from the acute to chronic period following mTBI using advanced time-frequency event-related potential (ERP) analysis, which isolates rapid regional brain activity and measures the functional communication within and between brain networks in response to varying task stimuli. The validity of these ERP biomarkers was evaluated with correlations between abnormal ERP findings and widely used clinical measures of cognitive functioning (i.e., neuropsychological tests and self-reported cognitive symptoms). Differences between mTBI caused by blast explosion versus impact to the head were also evaluated. A sample of 173 service members, comprising an mTBI group, an orthopedically-injured control group, and a healthy control group, completed ERP, neuropsychological, and self-report assessments within weeks following injury and again six months later. Results suggested that mTBI leads to cognitive changes that persist in the acute to post-acute period following injury (i.e., up to 12 weeks). These cognitive changes were reflected by alterations in ERP time-frequency amplitude and functional connectivity measures, and they remained apparent even when controlling for psychiatric symptoms. ERP differences were also evident between blast-related and impact-related mTBI. Conversely, neuropsychological test performance was not sensitive to mTBI. Abnormal ERP time-frequency measures were related to self-reported cognitive symptoms, suggesting these ERP measures are valid biomarkers of cognitive difficulties following mTBI. Critically, cognitive functioning as assessed by ERP measures returned to a level indistinguishable from controls 7-9 months following mTBI, even though more than a third of mTBI patients continued to report cognitive symptoms. These persistent cognitive complaints were more related to post-injury psychiatric symptoms than to the direct effects of brain injury.