We studied the relationship between preoperative umbilical artery Doppler waveforms and umbilical vein pO2 and pH at elective cesarean section. An absence of end-diastolic velocities had a strong statistical association with hypoxia and acidosis, and was an accurate clinical test for hypoxia (sensitivity 78%, specificity 98%, positive predictive value 88%, and negative predictive value 98%) and acidosis (sensitivity 90%, specificity 92%, positive predictive value 53%, and negative predictive value 100%). It was also a clinically sensitive indicator of perinatal morbidity and mortality. Most fetuses with no end-diastolic velocities were growth-retarded, but the reverse was not true. The absence of end-diastolic velocities also divided both mature and immature fetuses into high- and low-risk groups for hypoxia and acidosis. In the presence of end-diastolic velocities, only very high S/D ratios (above 4.5) have any association with hypoxia. As a noninvasive test of fetal umbilical vein pO2 and pH, umbilical artery Doppler performs well.