We launched a prospective cranial ultrasound screening study at the Maternity Hospital of the University of Giessen to examine the incidence and severity of brain damage in newborns. More than 90% of all neonates born during 1984-86 were included in the study (n = 2781) and were screened for cerebral abnormalities on 5.9 +/- 3.3 (SD) day post partum. There were in 7.8% sonographic abnormalities, the most frequent being peri/intraventricular haemorrhages (5.2% PIVH of various degrees (grade I-III). Periventricular leucomalacia, porencephalia, subarachnoidal haemorrhages, and hydrocephali were rare (less than or equal to 0.2%). The incidence of PIVH increased progressively with decreasing gestational age, e.g. from 1.5% at 41 weeks up to 61% at 24-30 weeks of gestation. A large percentage of babies with cerebral haemorrhages were clinically normal. There was a close inverse relationship between the Apgar score at 1,5 and 10 min and both incidence and severity of PIVH. However, the relation between PIVH and both cardiotocography (matched-pairs analysis) and arterial cord blood-pH was poor. The incidence of PIVH was increased in growth retarded newborns (pH less than or equal to 7.29), preterm multiples and amnionitis, but not after preeclampsia and premature rupture of membranes. It is interesting to note that, in mature newborns (greater than or equal to 38 weeks), there was no difference in the incidence of PIVH between spontaneous deliveries, vacuum extractions, and Caesarean sections, and that there was no difference between vaginally delivered (1.5%) and sectioned breech presentations (3.5%) in this group. In preterms at 35-37 weeks with protracted labour and secondary Caesarean section, the incidence of PIVH was high (19%). We conclude, that neonatal cranial ultrasound screening is an important tool for risk assessment and quality control in obstetrics. It clearly improves both the quality of the advice given to the patients and the clinical management of high-risk pregnancies.