OBJECTIVE: The estimation of the fetal weight is of major interest in many situations when the route of delivery has to be determined: breech presentations, diabetes and suspected macrosomia. The objective was to evaluate the accuracy of different methods for estimating fetal weight (EFW) during labour. A quality audit of the ultrasound images was also performed. PATIENTS AND METHODS: This was a prospective study performed in thirty patients at 37 weeks gestation or more, during the second stage of labour. The accuracy of clinical versus ultrasonographic (US) approach for EFW was tested (by two formulas according to Hadlock, with one that does not account head circumference [HC]). The ultrasound images' quality was also evaluated. RESULTS: The clinical assessment was significantly more accurate (-1±8.5 %) than the US approach using the formula with HC (-8±9.8 %) (P<0.01). This difference was not found when choosing the formula without HC. Analysis of ultrasound images for abdomen and femur showed 74% and 89% of satisfactory measurements respectively, but only 43% for HC. CONCLUSION: These results did not demonstrate a superiority of the US over the clinical approach for EFW during labour. Taking into account the cephalic measurements appeared as a limitation of the accuracy of US formulas.