The outcome at a mean age of 8.3 years was determined for 88 of 89 surviving children with birth weights of 500 to 999 gm who were born in Victoria in 1979 and 1980; the condition of the untraced child, severely disabled at 5 years of age, was assumed to be unchanged at 8 years. Principal impairments in the children at 8 years of age were cerebral palsy 8 (9%); bilateral blindness 6 (6.7%) and poor vision in another 4 (4.5%); bilateral deafness requiring amplification 5 (5.6%); and epilepsy 2 (2.2%). The Full Scale score of the Weschler Intelligence Scales for Children--Revised was available for 77 children: 5 scores (6.5%) were less than 71 (low IQ), and 9 (11.7%) were between 71 and 85 (borderline IQ). Severe disability in 16 children (18%) was due to severe cerebral palsy, bilateral blindness, or low IQ. Mild or moderate disability was present in another 18 children (20.2%) with borderline IQ, mild or moderate cerebral palsy, poor vision, deafness, or epilepsy. Significantly more outborn children (10/18, 55.6%) than inborn children (24/71, 33.8%) were disabled. Ten children required special schooling; reading accuracy or comprehension was at least 18 months behind chronologic age in 17 (23%) of 75 children in normal school who were tested. The 2-year assessment correctly estimated the category of disability at 8 years of age in 55 (62%), overestimated it in 24 (27%), and underestimated it in 10 (11%). For the 82 children also seen at 5 years age, the category of disability at 8 years of age was correctly identified in 69 children (84%), overestimated in 4 (5%), and underestimated in 9 (11%). Change in psychologic test scores was the principal reason for reclassifying children. We conclude that the 2-yar assessment was valuable in the early identification of most of the severely disabled children, but developmental delay at 2 years of age did not always portend intellectual impairment at 8 years of age.