The recently introduced intrauterine growth curve, based on ultrasonically estimated foetal weights, was retrospectively applied to an inborn population of 883 infants born before 33 gestational weeks at the University Hospital of Lund, during 1985-94. The estimation of birthweight deviation resulted in 630 (71.3%) infants with a birthweight appropriate for gestational age (AGA), 244 (27.6%) infants with a birthweight small for gestational age (SGA) and 9 (1.1%) infants with a birthweight large for gestational age. Birthweight deviation was associated with an increased mortality [odds ratio (OR) adjusted for gestational age 1.29 per SD (12%) change in birthweight for gestational age, 95% CI: 1.10-1.50; p = 0.002]. At gestational age 25-28 weeks, SGA-infants had an increased incidence of respiratory distress syndrome (RDS) as compared to AGA-infants (OR adjusted for gestational age: 1.98, 95% CI: 1.12-3.52; p = 0.019). At gestational age 29-32 weeks, SGA-infants had a lower incidence of RDS as compared to AGA-infants (OR adjusted for gestational age: OR 0.52, 95% CI: 0.34-0.80; p = 0.003). After adjustment for confounding variables, infants born at gestational age 25-28 weeks from mothers with pre-eclampsia, appeared to be a high-risk group for RDS, whereas at the age of 29-32 gestational weeks, negative birthweight deviation had a protective effect against RDS. Antenatal corticosteroid administration appeared to have a less beneficial effect on mortality, RDS and cerebral haemorrhage in infants born SGA vs in those born AGA.