Ten years after the first seroprevalence study performed in Flanders, the aim of this cross sectional study was to follow the evolution of hepatitis A, B and C prevalence. The prevalence of hepatitis A antibodies, hepatitis B surface antigen and hepatitis C antibodies was measured in oral fluid samples collected by postal survey. Using the National Population Register, an incremental sampling plan was developed to obtain a representative sampling of the general population. A total of 24,000 persons were selected and 6,000 persons among them contacted in a first wave. With 1834 participants a response rate of 30.6% was achieved. The prevalence was weighted for age and was 20.2% (95% CI 19.43-21.08) for hepatitis A, 0.66% (95% CI 0.51-0.84) for hepatitis B surface antigen and 0.12% (95% CI 0.09-0.39) for hepatitis C. The prevalence of hepatitis A and C in the Flemish population is lower in 2003 compared with the results of the study performed in 1993. The difference may be due to a real decrease of the diseases but also to differences in the methodology. The prevalence of hepatitis B surface antigen remains stable. Considering the 30% response rate and the high quality of the self-collected samples as reflect of a good participation of the general population, saliva test for prevalence study is a good epidemiological monitoring tool.