Preterm birth is a major cause of neurological impairment in childhood and can lead to long-term clinical, educational and social problems [1]. Better understanding and measurement of cerebral function in infancy, would enable the functional consequences of brain damage to be identified at an earlier age. Cortical visual processing is an early-developing aspect of brain function and is essential for the acquisition of cognitive, motor and social skills. We have devised a test battery for examining functional vision, the ABCDEFV [3] which includes two tests of cortical function for young infants: (i) orientation-reversal visual event-related potentials (OR-VERP); (ii) latency of saccadic fixation shifts (FS) under competition, a test of cortically based control of attention. Other studies have shown that development of the OR-VERP in term infants with hypoxic-ischaemic brain insult correlates with neonatal neuroimaging [4] and along with FS performance is predictive of later developmental outcome [5]. In this study, we examined these indicators in a group of 24 prematurely born infants (<32 weeks gestation) who had neonatal and term MRI scans to test whether these measures indicated early brain damage and predicted later cognitive outcome. Infants were tested at between 3 and 7 months post-term. The OR-VERP was elicited with grating patterns switching between 45-135 deg orientations at 8 reversals s-1. The pass criterion was a statistically reliable OR-VERP (circular variance test) by 7 months post-term. In FS, the pass criterion was 80% correct refixations, and a difference between mean latencies for competition and non-competition trials of less than 0.5s. The Griffiths mental developmental scale was used as an outcome measure at 2 years. Infants were divided into three groups, Mild/Normal (n = 8), Moderate (n = 7) and Severe (n = 9), according to extent of white matter damage seen on T2 images. The majority of infants in the mild lesion group (7/8) showed normal OR-VERP responses, but only 3/7 of the moderate group and 2/9 of the severe group met the pass criterion. On FS, 6/8 of the infants in the mild group showed typical attention responses by 7 months but less than half the moderate group and none of the severe group passed the test. Absence of an OR-VERP response before 7 months predicted a DQ < 80 at 2 years with a sensitivity of 86% and a specificity of 65%. Failure on the FS test was also associated with a low DQ score (sensitivity 100%; specificity 53%). These results demonstrate that early visual cortical measures are useful diagnostic and prognostic tools to assess infants born prematurely. These measures provide not only a sensitive indicator of cortical visual function but also prediction of development across cognitive and motor domains.