We conducted a study to investigate the effect of parity on the following six serum markers used in screening for Down's syndrome, after adjusting them for ethnic group and maternal weight: alpha-fetoprotein (AFP), unconjugated oestriol (uE,), total human chorionic gonadotrophin (hCG), free a-hCG, free P-hCG, and dimeric inhibin A. We aimed to estimate the effect of adjusting for any differences found on the screening performance. AFP, uE,, and hCG concentrations were available from 16 666 women with singleton pregnancies without Down's syndrome or neural tube defects and without insulin-dependent diabetes mellitus, who were screened between 15 and 22 weeks' gestational age. Stored serum samples were available on a subset of 1347 women and these were used to measure free a-hCG, free B-hCG, and inhibin A. Serum concentrations were expressed as multiples of the median (MOM) for women of the same gestational age, weight, and ethnic group. Of the six markers, only hCG levels were affected by parity; hCG levels decreased by 3.1 per cent per previous birth (95 per cent confidence interval 2.24.0 per cent); there was no significant relationship between the number of previous abortions and hCG level after adjustment for the number of previous births. The effect of previous births on hCG was not due to maternal age. Only AFP was affected by maternal age, but the effect was small; levels increased by 4.4 per cent per 10 years of age (3-2-5.7 per cent). It is not worthwhile adjusting serum markers for parity or for maternal age in prenatal screening for Down's syndrome because their effect on the performance of screening is negligible.