AIM: To study the detection rates of second trimester ultrasound screening for neural tube defects (NTD), abdominal wall defects (AWD) and Down's syndrome (DS) in low risk populations at tertiary centers, and to compare the ultrasound screening detection rates with those that were obtainable by biochemical serum screening (double test: alpha-fetoprotein/human chorion gonadotrophin/age test). STUDY DESIGN: Prospective multicenter study with a three year inclusion period: 1/1/1989-31/12/1991. SUBJECTS: 27,844 low-risk women at 18-34 years of age who had a second trimester ultrasound screening examination. Of these, 10,264 also had a serum test. METHODS: An ultrasound malformation scan and a serum test were carried out at 17-19 weeks of gestation. Risk calculations regarding DS were based on alpha-fetoprotein, human chorion gonadotrophin and maternal age; performed retrospectively for the first two years. RESULTS: In total 73 cases were identified in the study population: NTD (n=34), AWD (n=7) and DS (n=32). The detection rates, (%, with 95% confidence interval) for ultrasound screening were: NTD: 79.4 (62.1-91.3); AWD: 85.7 (42.1-99.6); DS: 6.3 (0.8-20.8). In the subgroup of women who had both tests, the detection rates for ultrasound screening vs double test were: NTD: 62.5 (24.5-91.5) vs 75.0 (34.9-96.8); AWD: 66.7 (9.4-99.2) vs 100 (29.2-100.0); DS: 7.7 (0.2-36.0) vs 46.2 (19.2-74.9). The false positive rates (%) for ultrasound screening vs double test were: NTD: 0.01/3.3; AWD: 0.01/3.3; DS: 0.1/4.0. CONCLUSION: Second trimester ultrasound screening in a low risk population gave a low detection rate for fetal DS (6.3%) and an acceptable detection rate for NTD (79.4%) and AWD (85.7%). In the subgroup of women who had both tests, serum screening performed better than ultrasound as applied in the present study, especially regarding DS.