ABSTRACTObjectiveTo investigate the association between large‐for‐gestational‐age (LGA) pregnancy and stillbirth to determine if the LGA fetus may benefit from antenatal testing with non‐stress test or biophysical profile.MethodsThis was a retrospective cohort study of singleton pregnancies that were ongoing at 24 weeks' gestation and that had undergone routine second‐trimester anatomy ultrasound examination, during the period 1990 to 2009. Pregnancies complicated by fetal anomaly or aneuploidy, those with missing birth weight information and those that were small‐for‐gestational age were excluded. Appropriate‐for‐gestational age (AGA) and LGA were defined as birth weight between the 10th and 90th percentiles and > 90th percentile, respectively, according to the Alexander growth standard. The incidence of stillbirth was calculated as the number of stillbirths per 10 000 ongoing pregnancies. Adjusted odds ratios (aOR) with 95% CI for stillbirth in LGA compared with AGA pregnancies were estimated using logistic regression analysis, controlling for pre‐existing and gestational diabetes. The incidence and aOR for stillbirth were estimated at 4‐week intervals from ≥ 24 to ≥ 40 weeks' gestation.ResultsOf 52 749 pregnancies ongoing at 24 weeks, 46 205 (87.6%) were AGA and 6544 (12.4%) were LGA at delivery. The incidence of stillbirth in LGA pregnancies was significantly higher than that in AGA pregnancies from 36 weeks' gestation (26/10 000 vs 7/10 000; aOR, 3.10; 95% CI, 1.68–5.70). When women with diabetes were excluded in stratified analysis, pregnancies complicated by LGA continued to be at increased risk for stillbirth ≥ 36 weeks (18/10 000 vs 7/10 000; OR, 2.63; 95% CI, 1.27–5.43).ConclusionPregnancies complicated by LGA are at significantly increased risk for stillbirth at or beyond 36 weeks, independent of maternal diabetes status, and may benefit from antenatal testing. Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.