Many clinicians use ultrasound estimates of fetal weight to assess fetal growth. This study assessed the accuracy of the Hadlock IV equation, the equation used at the Royal Hospital for Women, Randwick, NSW, in estimating birth weight. The accuracy of the Hadlock IV equation was assessed based on systematic and random error as well as absolute error. 709 women who underwent ultrasound examination within 8 weeks of delivery between January 2009–May 2011 were included. 305 women underwent ultrasound less than 2 weeks before delivery. The systematic, random and absolute errors of the Hadlock IV equation were −0.47, 27.45 and 8.52%, respectively. This study demonstrates that clinicians may rely on ultrasound estimates of fetal weight performed by well‐trained staff in a tertiary institution within 2 weeks of delivery. The accuracy diminishes as the interval between testing and delivery increases. Absolute errors tended to increase with increasing birth weight for all ultrasound‐delivery intervals.