As part of a prospective follow-up study of two premature cohorts (gestational age less than or equal to 36 weeks) born between 1976 and 1977, designed to evaluate the effects of IUGR on morbidity, mortality and neurological development in prematurely born children, 71 prematurely born children (48 AGA, 23 IUGR) without major neurological handicaps were subjected to detailed assessment of their neurological and psychological status at the age of 4 years. Thirty-six healthy full-term children formed a control group. The socioeconomic status of the families of the premature groups was similar to that of the families of the control group. The assessment consisted of a neurodevelopmental screening test and four psychological tests. Validation of the NDS test in relation to proven cognitive problems (sensitivity 96%, specificity 64%, relative risk 16.74) and determination of normal ranges were based on findings in the control group. There were significantly more children with moderate to high risk scores (greater than or equal to 10) among the IUGR group than among the AGA group. The neurodevelopmental profile of the preterm IUGR group was characterized by complex deviations of motor, visual and perceptual functions from normal. Preterm AGA children had only slightly higher risk scores in relation to fine motor, upper motor and perceptive functions than control children. Of the various perinatal factors studied, IUGR (especially in relation to boys who needed respirator therapy) was the most likely to be associated with abnormal NDS scores before school age.