The objective of this study was to test the hypothesis that neurodevelopmental outcome would differ between two groups of small-for-gestational age infants born at term showing different in utero growth characteristics during the third trimester. The design was a prospective cohort study done at a tertiary referral centre. The 76 subjects who fulfilled the inclusion criteria had an estimated fetal weight below the 10th centile for their gestation. Subsequent to enrolment, repeat ultrasound scans were performed weekly to determine growth velocity. Twenty-three infants whose change in fetal abdominal circumference between first and last scan was greater than -1.5 standard deviation scores (SDS) were assigned to the intrauterine growth retardation group (IUGR) while 53 infants whose fetal abdominal circumference changed less than 1.5 SDS were assigned to the small-for-gestational-age (SGA) group. Ten infants with normal intrauterine growth were enrolled as controls. Following delivery all infants had a neurological examination and a cranial ultrasound scan. At 1 year, 75 infants (87%) were traced and reassessed (49 SGA, 18 IUGR and eight controls) with a neurological examination and a developmental assessment. At birth, impairments were found in 27 (51%) of the SGA, 13 (57%) of the IUGR groups and one (10%) of the controls. At 1 year, 18 (37%) of the SGA subjects, six (33%) of the IUGR subjects and one (13%) of the control infant were impaired, including three (6%) of the SGA subjects and one (6%) of the IUGR subjects who were disabled. We conclude that in term fetuses with an estimated birthweight below the 10th centile for their gestation, the pattern of growth in the third trimester does not affect outcome at 1 year. In spite of optimum obstetric management, nearly one-third of the combined SGA and IUGR term fetuses had suffered some, albeit minor, neurological damage.