AIMS: To validate the sensitivity of universal antenatal screening for hepatitis B surface antigen (HBsAg) by testing pools of 10 sera, and to review 10 years' experience using this method. METHODS: 66,945 antenatal patients were tested between 1986 and 1996 using the pooled method. All sera from 1996 (n = 6050) were retrieved and retrospectively tested individually. An in vitro determination of the effect of pooling on sensitivity was performed by checkerboard neutralisation assay. RESULTS: 26 HBsAg positive women were detected by universal screening over 10 years; 12 had non-European surnames and five had known risk factors for hepatitis B infection. High titre anti-HBs sera in the pool reduced the sensitivity of the HBsAg assay, though the effect was only significant at low levels of HBsAg carriage. CONCLUSIONS: The prevalence of hepatitis B is extremely low in the antenatal population served by Plymouth PHL. Pooling is unlikely to reduce sensitivity enough to lead to significant preventable vertical transmission, and is a cost-effective and valid strategy in areas of low seroprevalence.