<p>Mild traumatic brain injury (mTBI) is among the most common neurological presentations worldwide, yet its diagnosis remains largely clinical and its prognosis difficult to predict. Structural imaging is typically normal and self-reported symptom scales are insensitive and confounded by effort and comorbidity. Because the generation and control of eye movements recruit an unusually large and anatomically distributed proportion of the brain—from brainstem and cerebellar circuits to parietal, frontal and prefrontal cortices linked by long white-matter tracts—ocular-motor behaviour is sensitive to the diffuse axonal and neurometabolic disturbance that characterises mTBI. This has motivated intense interest in eye movements as objective, quantifiable and increasingly portable biomarkers of injury and recovery. This review examines the anatomical susceptibility, measurement and current evidence base for the principal ocular domains affected in mTBI: saccades, smooth pursuit, fixation, vergence, the vestibulo-ocular reflex, the pupillary light reflex and accommodation. Across domains the literature converges on a recurring pattern in which cognitively demanding tasks reveal deficits that simple reflexive paradigms miss, supporting the view that ocular-motor dysfunction in mTBI is, in large part, a readout of distributed cognitive control rather than focal damage to ocular-motor machinery. We appraise diagnostic accuracy and the relationship between ocular metrics, symptom burden, white-matter integrity and recovery, and we critically address the methodological heterogeneity that currently prevents any single ocular metric from achieving validated-biomarker status. We argue that the most promising path is a standardised, multi-domain composite, validated longitudinally against converging neuroimaging and clinical endpoints, and we set out the methodological agenda required to bring quantitative eye-tracking into routine clinical practice.</p>