Some research suggests the National Adult Reading Test (NART) may be impaired by traumatic brain injury (TBI). To investigate this, a prospective, longitudinal study included 214 Australian TBI participants given the NART within 1 month post-injury, and at 6 and 12 months. TBI severity, age, education, sex, and mood (HADS) were examined, and significant improvement in NART- estimated full-scale IQ (FSIQ) was noted over time (p < .001). A three-way interaction of time, severity, and age showed younger and middleaged mild TBI sub-groups improved significantly between 1 and 6 months post-injury, and the older sub-group between 6 and 12 months. In severe TBI, significant NART-estimated FSIQ improvement was noted only for the middle-aged and only between 6 and 12 months post-TBI. NART administration soon after TBI underestimated premorbid IQ, with a complex relationship between time post-injury, severity, and age. Participants with a longer education had higher NART FSIQs, although those with less education gained more IQ points between initial and 6 months assessments suggesting higher initial NART impairment. Regression equations predicted NART FSIQ at 6 (R = 0.813) and 12 months post-trauma (R = 0.792-0.845), their effectiveness being confirmed via cross-validation. No significant relationships were noted between NART-estimated FSIQ and mood variables, or evidence of significant gains in estimated FSIQ between 12 and 24 months post-injury.