Objective: To investigate possible delivery-related impaired neonatal hepatocellular integrity by assessment of arterial and venous umbilical cord plasma levels of glutathione S-transferase Alpha 1-1. Methods: Glutathione S-transferase Alpha 1-1 levels were assessed in arterial and venous umbilical cord, and maternal venous plasma samples. The influence of maternal, delivery, and neonatal characteristics on arterial umbilical cord glutathione S-transferase Alpha 1-1 levels was studied, using linear regression analysis after log-transformation. Results: Median (range) arterial umbilical cord glutathione S-transferase Alpha 1-1 plasma levels were higher than venous umbilical cord levels (9.68 [0.64-1125] mu g/L and 7.66 [0.78-987.5] mu g/L, respectively, P < .005). Median (range) arterial and venous umbilical cord glutathione S-transferase Alpha 1-1 levels were higher than, and did not correlate with, maternal venous plasma levels (8.79 [1.79-183] mu g/L and 6.47 [1.58-164.5] mu g/L versus 1.47 [0.46-10.4] mu g/L, P < .001). Neonates born vaginally demonstrated higher median (range) levels than those delivered by cesarean (13.41 [1.02-1125] mu g/L and 5.73 [0.64-172.90] mu g/L, respectively, P < .001). Neonates with unfavorable pH (arterial pH under 7.20) demonstrated higher median (range) levels than those with normal pH (arterial pH at least 7.20) (15.15 [0.77-1125] mu g/L and 8.82 [0.64-120.90] mu g/L, respectively, P < .001). Stepwise multiple linear regression analysis showed that birth weight had the largest influence on arterial umbilical cord glutathione S-transferase Alpha 1-1 levels, followed by arterial base deficit, and route of delivery. Conclusion: Arterial umbilical cord glutathione S-transferase Alpha 1-1 plasma levels, being unrelated to maternal venous levels, might give a reliable impression of early neonatal hepatocellular integrity and may become an additional indicator of neonatal condition immediately after birth. (C) 1998 by The American College of Obstetricians and Gynecologists.