Objective-Evaluation of prenatal diagnosis of neural tube defects by ultrasound examination in unselected populations across Europe. Setting-Prenatal ultrasound units in areas that report to contributing congenital malformation registers. Methods-All cases with a suspected or confirmed neural tube defect and delivered within the 30 month study period were identified from 18 Congenital Malformation Registers from 11 European countries. Data on the pregnancy, prenatal scans, outcome of pregnancy, and information on diVerent screening policies for each country were analysed. Results-670 766 deliveries occurred in the area covered by the registers during the study period. A neural tube defect was diagnosed at delivery in 542 cases. In 84% of these, the lesion was isolated (166 anencephaly, 252 spina bifida, 35 encephalocele). Of the 166 isolated cases with anencephaly, 96% were correctly identified prenatally; one was missed on scan, two were wrongly diagnosed, and four were not scanned (sensitivity 98%). 84% of the prenatal diagnoses were made before 24 weeks' gestation; 86% of isolated anencephalic pregnancies were terminated. Of the 252 cases of isolated spina bifida, 171 (68%) were correctly identified prenatally; 66% of these before 24 weeks' gestation. The diagnosis was missed on scan in 60 cases and 21 were not scanned (sensitivity 75%). The mean reduction in birth prevalence because of termination of pregnancy for spina bifida was 49% (range 6-100%). There was a wide variation between centres in prenatal detection rate (33-100%), termination of pregnancy of prenatally diagnosed cases (17-100%), and gestation both at diagnosis and termination of pregnancy. Conclusion-A high prenatal detection rate for anencephaly was reported by all registers. There is a large variation in prenatal detection and termination rates for spina bifida between centres, reflecting diVerences both in policy and culture. (J Med Screen 2000;7:169-174)