Traumatic brain injury (TBI) is the leading cause of death and disability in young adults in the developed world. After accounting for known prognostic factors, there remains marked unexplained heterogeneity in outcome between individuals with overtly similar injuries, and furthermore, around one third of patients go on to develop a chronic neurodegenerative process. There is currently no known explanation for either the unpredictability in outcome or chronic neurodegeneration, but neuroinflammation has been implicated as potentially contributory to both, and represents a tractable therapeutic target. Most research has focussed on the innate immune response to injury, but a few papers have described autoantibodies to brain proteins occurring following TBI. This thesis describes the development of a protein microarray approach for screening serum for autoantibodies following TBI, and uses this to investigate the autoantibody responses which occur after TBI, both during the acute phase and at long-term time points up to many years post-injury. Two distinct responses appear to occur contemporaneously: 1) a broad “polyantigenic” upregulation of antibody binding to the majority of antigens on the microarray, and 2) dominant autoantibody responses to particular antigens, most commonly Myelin Associated Glycoprotein (MAG). The temporal profile for both of these responses follows a canonical inoculation pattern, dominated by IgM in the acute phase, but switching predominantly to IgG at follow-up months after the injury. The dominant responses to particular antigens appear to wane when assessed years post-injury, but polyantigenic upregulation of IgG binding persists even at this long-term time point. Subacute polyantigenic IgM upregulation occurs most markedly in young adults, and appears to associate with a worse outcome than would be predicted using recognised prognostic factors. In addition, the persistence of anti-MAG IgM autoantibodies months to years after injury correlates with biomarkers of ongoing neurodegeneration. The use of enzyme-linked immunosorbent assays to measure autoantibodies in this setting is assessed, but their signal is dominated by the magnitude of polyantigenic response, and therefore cannot accurately differentiate dominant responses to particular antigens. Lastly, pilot data for future work investigating cellular immunological responses to TBI are appended; methodology is developed to isolate mononuclear cells from cerebrospinal fluid collected from external ventricular drains, with a view to investigating the trafficking of lymphocytes into the central nervous system where tertiary lymphoid structures might theoretically form and maintain autoantibody production inside the blood brain barrier.