Abstract Aims: To analyze the clinical differences between early- and late-onset cases of severe preeclampsia and to evaluate parameters that could help to predict perinatal outcome. Methods: Over a period of 6 years, all cases of severe preeclampsia (n=68) at our institution were included in a retrospective cohort analysis. Differences between early (<34 weeks, n=44) and late (≥34 weeks, n=24) onset of the disease were evaluated. Risk factors for low 5-min Apgar score (≤7), small-for-gestational-age (SGA) infants and neonatal acidosis (umbilical arterial pH <7.20) were identified and considered in a multiple logistic regression model. Results: Early- and late-onset severe preeclampsia differed from each other remarkably. Perinatal outcome was unfavorable in early-onset disease and seemed to be mainly a result of premature delivery and development of fetal growth restriction. Abnormal uterine Doppler velocimetry increased the risk of low 5-min Apgar values [odds ratio (OR) 8.0, P=0.012] and preterm birth <34+0 weeks (OR 17.9, P<0.001). An increased resistance of the umbilical artery was associated with a higher risk for SGA birth weight (OR 4.9, P=0.010). Conclusion: Preeclampsia is a heterogeneous syndrome even if only severe cases were analyzed. Abnormal Doppler flow characteristics facilitated the identification of patients who were at increased risk for worse perinatal outcome.