Objective. To test the hypothesis that the (case) twin later to succumb in utero has biparietal diameter measurements (by ultrasound) different from those (control) twins surviving the perinatal period.

Material and methods. Information from the Medical Birth Registry, National Board of Health and Welfare, Stockholm, was used to identify all births in a defined population in southern Sweden with about 20,000 deliveries each year. In 3,019 twin pregnancies between 1973 and 1989, one or both twins were stillborn in 47 cases (gestational duration ≥28 weeks, birthweight ≥500g.). For each case pregnancy, two control pregnancies were selected, the matching criteria being: same delivery unit, same parity (0, I, II, III+), similar year of delivery (± 1 year) and maternal age (± 5 years). Data on ultrasound examinations were extracted from the original medical records. Screening in early second trimester started at one of the units as early as 1973 and at the latest of 12 units in 1982.

Results. There was no obvious difference between cases and controls in intra-pair discordant biparietal diameter (BPD) measured in early second trimester. Nor was there any evident difference in the rate of deviant BPD between cases and controls. In all, 8% of dead male and 24% of dead female fetuses were by definition small-for-gestational age (<-2 standard deviations).

Conclusions. No significant difference was seen between cases and controls regarding deviating biparietal diameters. Abdominal diameter may be a better predictor of subsequent fetal death (not analysed in this study), though only about 15% of all dead twins were deemed small-for-gestational age.