Aims: In our study we determined possible risk factors for intraventricular hemorrhage grade III to IV (IVH) based on a regional German neonatal data base and tried to build a logistic-regression model to predict the risk of IVH according to gestational age. Materials: We identified 3721 premature infants, 22 to 36 completed weeks of gestational age, born from 1994 through 1997. 136 (3.7%) IVH were diagnosed sonographically. 60 (44%) infants with IVH died. We examined the following variables as risk factors for IVH: gestational age, sex, blood pH of 7.2 or less, body temperature of 35°C or less, multiple birth, small-for-gestational age, intubation after birth, transport to another hospital. Results: In the full logistic regression model sex, blood pH of 7.2 or less, multiple birth, and small-for-gestational age were not associated with a significant risk of IVH. Body temperature of 35°C or less was associated with an increased risk of IVH (adjusted odds ratio, 1.92; 95% confidence interval, 1.09 to 3.40). Intubation after birth increased the risk of IVH in neonates under 28 weeks of gestational age (OR, 3.72; 95% CI, 1,65 to 8.38) only to a moderate extent, but significantly in neonates 32 to 36 weeks of gestational age (OR, 16.51; 95% CI: 7.35 to 36.18). The risk of IVH was mainly related to gestational age. Neonates delivered before 28 weeks of gestation (OR, 75.72; 95% CI, 46.14 to 124.30) faced the highest risk of IVH. Transport to another hospital was connected with an increased risk of IVH regardless of gestational age (adjusted OR, 1.95; 95% CI, 1.07 to 2.56). Conclusion: The frequency of IVH could be reduced significantly, if extremely premature infants, the vast majority of patients suffering from IVH, did not have to be transferred postnatally to another hospital.