Fetal and maternal plasma noradrenaline responses to invasive procedures were determined in pregnancies of 18 to 37 wk gestation. Fetal umbilical venous blood sampling was performed either from the placental cord insertion, which is not innervated, or the intrahepatic vein, which is innervated, and thus may be more stressful for the fetus. Samples from diagnostic procedures, as well as from transfusion procedures, were compared between the two sites. Fetal plasma levels were significantly elevated in blood samples obtained from the intrahepatic vein compared with those from the placental cord insertion during diagnostic procedures [p < 0.05, geometric means and 95% confidence intervals (CI) were 0.67 nmol/L (0.43-1.04) and 0.36 nmol/L (0.25-0.54), respectively]. Plasma levels in samples taken before transfusion from the intrahepatic vein were also significantly higher than those from the placental cord insertion. After transfusion, there was a significant rise in fetal plasma noradrenaline levels at both sites; however, after transfusion through the intrahepatic vein, the rise was substantially greater than after transfusion through the placental cord insertion (p < 0.05, change, mean deltaNA, and 95% CI were 0.67 (0.37-1.22), and 0.20 (0.12-0.33), respectively). The deltaNA was significantly associated with the duration of the stimulus (the time the needle remained in situ) (p = 0.05, adjusted R2 = 0.48) and with gestational age. Maternal levels rose substantially and equally after transfusions at either site (mean deltaNA and 95% CI, 6.46 nmol/L, 1.74 to 11.18 and 9.49 nmol/L, 6.24 to 12.75 for the intrahepatic vein and placental cord insertion groups, respectively). There was no significant correlation between baseline fetal and maternal levels (r = 0.08, n = 41) or between deltaNA pre- and posttransfusion maternal and fetal values in either group. These results indicate that the fetus is capable of mounting an independent noradrenaline stress response to a needle transgressing its trunk from 18 wk gestation. The effect was observable in samples taken at a mean of 5.6 min after needling. The lack of correlation between maternal and fetal levels suggests that virtually no noradrenaline crosses the placenta directly, and that the observed fetal responses are not due to direct transport from the mother.