Two recent publications by Cuckle et al., and one each by Canick et al. and Kellner et al., describe the use of urine beta-core fragment measurements as a screening test for Down syndrome pregnancies. Median levels of over 5.4 MOM were reported for cases of Down syndrome, with an over 72 per cent detection rate for a 5 per cent false-positive rate. Urine beta-core fragment was suggested as a superior screening test for Down syndrome pregnancies. These four studies were retrospectives, with samples from affected cases collected at different sites from those from normal cases. In the present study, prospective data were collected for 726 pregnancies over a 9-month period at a single medical centre. Fresh samples were assayed continuously, without knowledge of the karyotype. Urinary beta-core fragment levels in 709 unaffected samples continually declined from 12 to 24 weeks of pregnancy. A logarithmic fit was optimal for the median curve. The log standard deviation of unaffected samples was 0.368. All 13 Down syndrome cases had levels exceeding 1.0 MOM, with a median value of 4.1 MOM. Eight of 13 Down syndrome cases (62 per cent) had levels exceeding the 95th centile. Results have not been adjusted for maternal age, which may improve the detection rate. The results reported here, while less impressive than those reported previously, confirm the usefulness of urine beta-core fragment as a screening test for Down syndrome. Because of the prospective nature of this study, the 62 per cent sensitivity suggested here might be more representative of the true performance of urinary beta-core fragment in clinical practice than the higher rates observed in previous studies. Results for this single urine test are similar to those for triple screen and other serum combination tests. Single analyte urine beta-core fragment tests, or beta-core fragment combination protocols, may eventually replace serum analytes in screening for Down syndrome pregnancies.