Analysis of the waveforms obtained from the umbilical artery using Doppler ultrasound has been claimed in a number of recent reports to be a highly specific method of identifying babies with intrauterine growth retardation. Because of the morbidity and mortality associated with this condition it was important to evaluate the validity of those claims before introducing the method into obstetric practise. The work in this thesis was carried out in four parts. Firstly the technique was found to be reproducible with small intra and inter-observer errors. Secondly, possible influencing factors were studied, and it was found that the fetal heart rate and fetal breathing movements have a significant effect on the waveform. Thirdly, a longitudinal study of 50 healthy women with normal pregnancies was performed to define the normal range of A/B ratios (the principle method of analysis of the waveforms was the A/B ratio: ratio of peak systolic to end-diastolic frequency). Finally, Doppler ultrasound was performed in 205 women with high risk pregnancies, and results were compared to the normal reference data. When high risk cases with abnormal Doppler results (A/B ratio > 95th centile) were compared with high risk cases with normal Doppler results (A/B ratio < 95th centile) the Doppler abnormal group were found to have a significantly increased incidence of preterm delivery, operative delivery, late fetal heart rate decelerations in labour, increased admission to the special care baby unit, and increased neonatal morbidity. Although an abnormal Doppler result was associated with small for gestational age in 61% of cases, only 41% of cases of small for gestational age had abnormal Doppler results. The conclusion from this thesis is that Doppler flow velocity waveform analysis, although lacking the previously claimed diagnostic precision of detecting intrauterine growth retardation, could contribute to fetal management by identifying babies at high risk of intrapartum complications.