Three hundred and five pregnant women referred consecutively from a single antenatal clinic for imaging ultrasound after 26 weeks' gestation had doppler velocimetry of umbilical artery, uterine artery and umbilical vein. Results of doppler studies were not available to clinicians until after delivery and clinical case records were examined postpartum for the development of pregnancy-induced hypertension (PIH) or gestational diabetes mellitus (DM) and for fetal outcome. Accurate data were available for 272 women, of whom 167 had a normal pregnancy, 61 had or developed PIH and 44 had or developed DM. None of the various doppler velocimetry measurements differed significantly amongst these 3 groups at any gestational stage. Abnormal values for doppler systolic: diastolic (A/B) ratio were found in only 7% of umbilical artery measurements, 6% of uterine artery measurements and none of umbilical vein measurements. Specificity of abnormal results for PIH or DM was high (95-97%) but sensitivity low (16-17%) and positive and negative predictive values were generally low. Two perinatal deaths occurred and umbilical artery A/B ratio was abnormal in both cases. This study shows that there is a low yield of abnormal results for doppler velocimetry of umbilical artery, uterine artery or umbilical vein in routine antenatal screening.