Introduction
Small‐for‐gestational‐age (SGA) confers a higher perinatal risk of adverse outcomes. Birthweight cannot be accurately measured until delivery, therefore accurate estimated fetal weight (EFW) based on ultrasonography is important in identifying this high‐risk population. We aimed to establish the sensitivity of detecting SGA infants antenatally in a unit with a selective third‐trimester ultrasound policy and to investigate the association between EFW and birthweight in these babies.

Material and methods
A retrospective cohort study was conducted on non‐anomalous singleton pregnancies delivered after 36 weeks of gestation where SGA (<10th percentile) was diagnosed at delivery. The EFW at the time of the third‐trimester ultrasound scan was recorded using standard Hadlock formulae.

Results
In 2017, there were 8392 non‐anomalous singleton pregnancies live born after 36 weeks, excluding late bookers. 797 were live‐born SGA <10th percentile for birthweight and 464 <5th percentile, who met our inclusion criteria. The antenatal detection rate of SGA was 19.6% for babies with birthweight <10th percentile and 24.1% <5th percentile. There was a significant correlation between the EFW and birthweight of fetuses undergoing ultrasound assessment within 2 weeks of delivery (P < .001, r = 0.73 (Pearson correlation). For these cases, EFW was greater than the birthweight in 65% of cases. After adjusting all EFWs using the discrepancy between EFW and actual birthweight for those babies born within 48 hours of the scan, the mean difference between the birthweight and adjusted EFW 7 days before delivery was 111 g (95% CI 87‐136 g) and at 14 days was 200 g (95% CI 153‐248 g). Despite adjusting the EFW, 61/213 cases (28.6%) apparently lost weight between the ultrasound scan and delivery.

Conclusions
Small‐for‐gestational‐age infants with a birthweight <10th percentile are poorly identified antenatally with little improvement for those <5th percentile. In SGA babies, ultrasound EFW overestimated birthweight. Discrepancies between birthweight and EFW are not explicable only by the limitations of third‐trimester sonography, a reduction in fetal weight close to delivery in a proportion of liveborn SGA babies is plausible.