The results of x-ray pelvimetry performed on 66 pregnant women, with a view to predicting the outcome of pregnancy, were reviewed. It was concluded that such pelvimetry was most valuable in cases in which it indicated a normal pelvis or gross bony disproportion. It was least effective in those with a "borderline" pelvis, in which the correct management of the patient requires an adequate trial of labour. The authors believe that pelvimetry in the later stages of pregnancy has no ill effects on the unborn baby or the mother.Indications for use of x-ray pelvimetry in obstetrics, the technique employed, the various types of female pelvis and the relation of these to the course of labour are considered. Pelvimetry findings are compared with the eventual outcome of labour in women with suspected cephalopelvic disproportion. It is emphasized that the pelvic assessment should be made jointly by the radiologist and obstetrician.