The objective of the study was to evaluate neonatal survival and subsequent disabilities in infants with extremely low gestational age in relation to perinatal events and neonatal treatment. A retrospective follow-up study was performed based on medical records, questionnaires to parents and recordings of contact with health authorities. All infants with a gestational age of 28 completed weeks or less, who were admitted to the Department of Neonatology, Rigshospitalet. within 24 hours of age during the period January 1, 1987 - December 31, 1990 were included. During this period the basic therapeutic approach was a combination of minimal handling and early nasal-continuous positive airway pressure (CPAP) (''minitouch''). Main outcome measures were: mortality, healthy survival and disabled survival. Variables related to outcome were: risk factors present at birth (gestational age, birth weight, gender, place of birth (Rigshospitalet/other hospital), mode of delivery, Apgar score at five minutes: interventions in the neonatal period (intermittent positive pressure ventilation and treatment of hypotension): complications in the neonatal period (intracranial haemorrhage grade II-IV, periventricular leucomalacia, pneumothorax, seizures and septicaemia). One hundred and ninety-seven infants without major malformations were included. The mortality rate was 29%. Among infants with gestational age 24-25 weeks 49% died versus 24% of infants born after 26-28 weeks (p=0.004). Mean gestational age was 26.7 weeks (range 24-28) and mean birth weight 994 g (range 525-1630). Fifty-five infants (28%) were small-for-gestational age. One hundred and fifty-five infants (79%) were born in our hospital and 115 (58%) were delivered by caesarean section. A total of 140 infants (71%) survived until discharge and none died between discharge and follow-up At follow-up at a mean uncorrected age of 48 months information was obtained about all infants, except two (1%) who had emigrated: 75 (54%) had no impairments, 31 (22%) had minor impairments, 17 (12%) were moderately disabled, and 15 (11%) were severely disabled. Of the 197 infants 121 (61%) were treated with intermittent positive pressure ventilation. 83 (42%) with dopamine for hypotension, and 92 (47%) received parenteral nutrition. In 64 infants (33%) the course was complicated with intracranial haemorrhage (ICH) grade II-IV, in 17 (9%) with seizures, in 23 (12%) with pneumothorax, in 21 (11%) with septicaemia, and in 10 (5%) with necrotizing enterocolitis. Sixty infants (31%) needed medical or surgical closure of a persistent ductus arteriosus. In 11 infants (6%) cystic periventricular leuco-malacia occurred, 10 infants (5%) developed retinopathy of prematurity stage 3-4, and 35 infants (24%) received supplementary oxygen at 28 days of age. Risk factors present at birth for adverse outcome were: Apgar score <7 at five minutes, birth weight <1000 g, male sex, and birth in another hospital than Rigshospitalet. For adverse outcome in surviving infants only, ICH grade II-IV was the only significant risk factor.