Sera from an age-stratified sample of 4188 individuals, submitted for diagnostic purposes to 15 public health laboratories in England and Wales in 1996, were tested for hepatitis A antibody. The serological profiles were consistent with declining incidence in the past. This hypothesis was tested by comparing the serological profiles of Ashford, Leeds and Preston public health laboratories with those from sera collected during a previous study in the same laboratories in 1986/7. A comparison of equivalent 10 year birth cohorts revealed that significant hepatitis A seroconversion had only continued in Ashford. However, it is probable that most seroconversions are due to vaccination and immigration rather than continuing viral transmission. Further population-based surveys collecting more in-depth social and demographic data are needed to confirm the main factors influencing hepatitis A seroprevalence and to explain the regional differences.