The relationship between the Glasgow Coma Scale (GCS) and neuropsychiatric outcome was examined in 57 consecutive subjects with mild traumatic brain injury (TBI) attending a follow-up clinic. Subjects were grouped according to initial GCS score (15 versus 13-14) and contrasted at an average of 5-6 months post-injury. As expected, those with GCS 13-14 had longer PTA (p = 0.001) and a higher rate of abnormal brain CT scans (p = 0.005). However, no significant differences emerged for indices of neuropsychiatric status, including measures of neurobehavioural symptoms/signs, overall psychological distress, psychiatric 'caseness', functional and psychosocial outcome, frequency of common somatic complaints, and rate of return to work. Subsidiary analyses based upon the presence/absence of CT abnormalities and the duration of PTA (<1 hour versus 1-24 hours) also failed to predict outcome, although a trend associating longer PTA with lower functional outcome was observed. Thus, despite early neurosurgical differences, the results suggest that initial GCS scores do not clearly translate into neuropsychiatric sequelae at follow-up within the rubric of GCS 13-15.