Objective To determine the prevalence of increased fetal nuchal translucency thickness in twin pregnancies and to evaluate screening for trisomy 21 by a combination of translucency thickness and maternal age.

Design Prospective screening study at 10 to 14 weeks of gestation.

Setting Fetal Medicine Centre.

Population 22,518 self-selected pregnant women at 10 to 14 weeks of gestation, including 21,477 singleton and 448 twin pregnancies with live fetuses.

Methods Fetal nuchal translucency thickness was measured by ultrasound examination at 10–14 weeks. Sensitivity and false positive rates of screening for trisomy 21 by a combination of fetal nuchal translucency thickness and maternal age were calculated.

Main outcome measures Prevalence of increased nuchal translucency thickness and detection of trisomy 21.

Results In the 448 twin pregnancies the nuchal translucency thickness was above the 95th centile of the normal range (for crown-rump length in singletons) in 65/896 fetuses (7.3%), including 7/8 (88%) with trisomy 21. Increased translucency was also present in four fetuses with other chromosomal abnormalities. In the chromosomally normal twin prebmancies the prevalence of increased nuchal translucency was higher in fetuses from monochorionic (8.4%; 16/190) than in those with dichori-onic pregnancies (5.4%; 37/688). The minimum estimated risk for trisomy 21, based on maternal age and fetal nuchal translucency thickness, was 1 in 300 in 19.5% (175/896) of the twins including all eight of those with trisomy 21.

Conclusions In twin pregnancies the sensitivity of fetal nuchal translucency thickness in screening for trisomy 21 is similar to that in singleton pregnancies, but the specificity is lower because translucency is also increased in chromosomally normal monochorionic twin pregnancies.