Background and objective. Prenatal maternal serum screening for Down's syndrome has become an important and established part of modern antenatal care. Previously it has been reported that non-specific immunoreactive inhibin may be useful in this context. Using a novel assay we have evaluated dimeric inhibin A as a possible second trimester marker of Down's syndrome. Methods. From 1992-1993 records, stored sera from women with Down's affected pregnancies and chromosomally normal control pregnancies were identified and retrieved for analysis. These sera had been prospectively collected at 15, 16 and 17 weeks gestation. Subjects. Records revealed 21 women who had had a Down's syndrome pregnancy and who also had serum available for analysis. Sera from 150 chromosomally normal controls, matched for gestation and duration of storage, were also retrieved. Measurements. Dimeric inhibin A was measured using a recently developed two-site enzyme-linked immunoassay. This employs a capture anti inhibin β(A)-subunit monoclonal antibody, covalently bound to a microtitre plate and a second anti inhibin α-subunit antibody conjugated to alkaline phosphatase, allowing detection. Results. The mean (95% CI) maternal serum dimeric inhibin A in the samples from control pregnancies was 237 (201.5-273.4) ng/l, 266.9 (235.4-298.5) ng/l and 207.2 (178.5-235.9) ng/l at 15, 16 and 17 weeks gestation respectively. Expressing the results from the Down's samples as multiples of the normal median (MoM), the median (95% CI) MoM was 2.6 (2.25-3.57), significantly higher than the controls (P < 0.0001, Mann-Whitney U-test). In the sample set tested, for a given false positive rate of 5.3% inhibin A alone afforded a detection rate of 62%, detecting cases previously undetected by routine screening. Conclusions. Dimeric inhibin A appears to be a promising new marker for the prenatal detection of Down's syndrome. Further prospective evaluation and assessment with other established markers would now be merited.