The survival prospects for infants of birthweight less than or equal to 1500 g born in recent years have improved. Evidence for a corresponding decrease in long-term morbidity of survivors is conflicting but recent reports from some centers indicate that high morbidity rates are occurring. Until additional satisfactory reports are available on the outcome of very low birthweight (VLBW) infants born after 1975, preferably from a community or region, uncertainty will continue. The outcome of three cohorts of VLBW infants, born in the Royal Women's Hospital, Melbourne between 1966 and 1978 is reported; more than 90% of each cohort were fully assessed, aged 2-8 years. There were 169 long-term survivors from the first cohort (1966-1970 births) and 72 from the second cohort (1973-74 births); survival rates were 37.1% and 37.3% respectively; however, for the 1977-78 births, there were 161 survivors, a significant increase to 68.3%. In the first cohort, 32.7% had one or more visual defects and 3.9% were blind but visual morbidity decreased progressively in cohorts 2 and 3; 3% of the second cohort and 1.2% of the third cohort were blind. There was a trend for a decrease in severe sensorineural deafness. Cerebral palsy increased progressively, respectively 2.6%, 4.5% and 11.9% in the first, second and third cohorts. There was a significant increase in the mean Mental Developmental Index of the Bayley Scales at the age of 2 years from 75.38 for the 64 children born in 1966-70 compared with 90.96 for 150 children in the 1977-78 cohort. Although there had been an increase in upper social class families in the more recent cohort, improvement in test scores was still highly significant when higher social classes (1-3 Congalton Scale) were excluded. However, there was no significant improvement in the 6 year psychological test scores of the first and second cohorts. There was a steady increase in occurrence of cerebral palsy. Significance associations in the 1977-78 cohort were found with only 2 perinatal variables (use of theophylline and necrotizing entercolitis). Furthermore, 17 (89.5%) of children had a five-minute Apgar score greater than 5 and 14 (73.7%) did not require ventilatory support: Prevention of cerebral palsy by selective treatment in the delivery room or nursery was not feasible for prediction of this condition was not possible from perinatal risk factors.