We have examined maternal urine concentrations of beta core, free beta human chorionic gonadotrophin (hCG), and total oestriol in 373 control pregnancies and 43 pregnancies affected by aneuploidy (including 22 cases of Down's syndrome) in an attempt to see if any of the analytes have a value in Down's syndrome screening between the tenth and 14th week of pregnancy. We have compared the performance of these analytes against nuchal translucency measurement combined with maternal serum free beta hCG at the same period of pregnancy. Our results show that levels of urine free beta hCG and beta core are increased in Down's syndrome with average multiple of the median levels of 1.81 and 2.91, respectively. Urine total oestriol was reduced (0.83) whilst maternal serum free beta hCG was increased (1.72). In trisomy 18 the levels of all analytes were reduced, although serum free beta hCG was the most discriminating. The spread of results in the control and the Down's group for urine beta core was more than three times than that for serum free beta hCG and with urine free beta hCG it was two times wider. In combination with maternal age, urine total oestriol had a 32 per cent detection rate at a fixed 5 per cent false-positive rate; urine beta core 34 per cent, urine free beta hCG 36 per cent, maternal serum free beta hCG 44 per cent, and nuchal translucency 82 per cent. In combination with nuchal translucency, urine total oestriol added an extra 1 per cent detection, urine beta core an extra 2 per cent, urine free beta hCG an extra 3 per cent, and serum free beta hCG an extra 5 per cent. It is unlikely that any of the urine markers will be of value in first-trimester screening. Optimal first-trimester screening programmes will rely for the foreseeable future on nuchal translucency, serum free beta hCG, and possibly pregnancy-associated plasma protein A.