[Determination of the expected day of delivery--ultrasound has not been shown to be more accurate than the calendar method].

It has been claimed that the expected day of delivery (EDD) determined on the basis of an ultrasound scan is more accurate than using a calendar method. The aim of this paper is to assess the evidence in support of this claim. A critical review of the methodologically best research revealed that among women with regular cycles and know first day of the last menstrual period, Naegele's rule predicted the EDD to be 3.3 days too early, on average, whereas prediction based on ultrasound scans was 2.0 days too late. The standard deviations of the two distributions were the same. After correction for the systematic biases, the two methods of prediction were thus equally precise. It is concluded that the EDD should be calculated by adding 283 days to the date of the last period rather than by adding 280 days. For women with regular cycles and known first day of the last menstrual period, ultrasound dating does not lead to a more precise prediction.