ObjectivesTo evaluate the accuracy of ultrasound estimates of fetal weight made by residents.MethodsA cross‐sectional study was conducted on consecutive pregnancies attended in an obstetric ward. All participants provided their consent to participate. Inclusion criteria were singleton pregnancies with ultrasound evaluation and delivery within 14 days. Congenital abnormalities and stillbirths were excluded. Ultrasound scans were performed by two residents, one observer per subject, following standardized methodology and supervised by senior staff. Fetal weight was calculated using the four‐parameter Hadlock 1985 formula. Birth weight was measured using a standard clinical scale. Bland–Altman and mixed‐effects analyses were performed.ResultsIn a 6‐month period, 95 gestations were recruited at 30–41 weeks, and birth weights ranged from 1200 to 4080 g. The bias between methods was −28 g (CI95%, −74 to 18) with 95% limits of agreement of ±440 g. The 95% deviation between methods was of ±16%. Parity had the most significant effect on the bias between methods; operator and maternal weight had no significant effect.ConclusionsSupervised measurement of fetal weight by residents demonstrated high accuracy. Parity bias may be associated with visceral fat deposition independent of maternal weight gain. A methodology for medical evaluation is presented.