The obscuring effect of the diploe in the skull for attempts to use ultrasonic energy to image the brain is not present in the fetus, neonate or infant prior to ossification of the skull. Initially, however, the fetal examination was largely confined to a single measurement of the size of the skull, usually in the bi-parietal plane, just prior to labour. In this way it became possible to identify immature fetuses where delivery should be delayed as long as possible, as well as cases of cephalo-pelvic disproportion where the large size of the fetal head would make vaginal delivery difficult. Measurements made earlier in pregnancy, in order to predict the gestational age and estimated date of delivery, would only be accurate if all fetal heads were the same size at every gestational age. Similarly, attempts to predict birthweight from the size of the fetal head at any given gestational age foundered both upon the uncertainty of determining the gestational age from the menstrual history and upon the variations in the size of fetuses for various gestational ages. By making serial measurements of fetal head size, the rate of fetal growth can be measured. In this way, it is possible to separate the normal fetus which appeared to be small for its gestational age because the gestational age had been erroneously calculated from the menstrual history, from two clinically important groups. In the first group, the fetus is small and grows abnormally slowly throughout the whole pregnancy because of some congenital defect; and in the second the rate of growth becomes slowed towards the end of pregnancy because the placenta is no longer able to supply the increased nutritional needs of the mature fetus. Where it is not possible to make serial measurements throughout pregnancy, they can often be restricted to two measurements. The first is made in mid-pregnancy and, if the size of the fetal head agrees with that predicted from the menstrual history, it is assumed that the fetus is not abnormally small due to some congenital defect. In such cases the only other examination that needs to be made is another measurement near term to identify cases of placental insufficiency. If the first measurement does not coincide with that expected from the menstrual history, then more frequent measurements should be made in order to separate those normal fetuses in which the gestational age was wrongly calculated from the menstrual history from those fetuses with growth retardation throughout pregnancy due to some abnormality.(ABSTRACT TRUNCATED AT 400 WORDS)