A retrospective study was performed over a 5-year period (1990-94) to evaluate the effectiveness of prenatal ultrasonography in terms of sensitivity, specificity, and predictive values in detecting fetal anomalies by comparing prenatal ultrasonic results with anomalies found in neonates and the perinatal outcome of anomalous fetuses. Minor congenital anomalies as listed and defined in the Eurocat Register were excluded. From a total of 5544 singletons, 4819 had at least one ultrasound scan (87%), of which 3004 at low risk and 1815 (38%) at high risk for anomalies had routine screening (RS) and indicated scanning (IS), respectively. A total of 136 fetuses were structurally abnormal (2.82%, RS and IS: 0.77% and 6.23%) and 200 major anomalies (RS and IS: 37 and 163) were recorded. The overall sensitivity of the ultrasound test was 78.7% (RS and IS: 34.8% and 87.6%, P < 0.01) for abnormal fetuses and 58.0% (RS and IS: 29.7% and 64.4%, P < 0.01) for anomalies. The overall specificity was 99.9% and the positive and negative predictive values were 97.3% and 99.4%, respectively; these values did not differ significantly between the two groups. The sensitivity of ultrasound for the detection of abnormal fetuses before 24 weeks was 22.8% (RS and IS: 13.0% and 24.8%) which was associated with a 61% (25/41) termination rate (RS and IS: 25% and 75.9%, P < 0.01) and a 24.4% (10/41) postnatal survival rate (RS and IS: 41.7% and 17.2%). The overall survival rate following pre- and postnatal correction of anomalies was 44.9% (RS and IS: 60.9% and 41.6%). For the detection of fetal anomalies anatomic ultrasound scanning is necessary during pregnancy, irrespective of pregnancy condition. Early detection of fetal anomalies could offer the option of pregnancy termination.