Ultrasonic assessment of fetal weight derived from multiple measurements of the fetus has become accepted in clinical practice as being useful in singleton pregnancies. Several different formulae for estimating fetal weight have been derived from differing measurements of the fetus, such as biparietal diameter, trunk circumference, and femur length. To date, there has been no attempt to evaluate such a technique in multiple pregnancy. This study aims to see whether the formulae derived for singleton pregnancy are applicable to twin pregnancies. Estimated fetal weight will be derived by mathematical modelling from ultrasonic measurements made within a week of delivery and the results compared with the actual birthweight to give an indication of how good such estimated fetal weights are. Factors to be considered in the analysis include whether there are differences between Twin I and Twin II, differences of presentation with twins and differences by gestation at delivery.