Background: Although population-based serosurveys offer an optimal measure of cumulative infection rates, they are seldom performed due to high cost and complex logistics. Use of participant self-collected oral fluid as a diagnostic specimen and mail for specimen delivery has the potential of generating reliable, population-representative data at limited cost. Methods: A survey of oral fluid HAV-specific immunoglobulin G (an indicator of past HAV infection) was undertaken in a provincially representative sample of 20-39 year olds as a pilot study. A provincial administrative database served as the sampling frame. Potential participants were invited by mail to collect oral fluid and complete a questionnaire at home and return both by mail. Additional telephone prompting was directed at slow responders. Oral fluid was tested using a validated ELISA. Results: From among 2,448 potential participants, contact by mail or telephone was made with 1,009 eligible subjects; 59% (585) participated. Materials withstood mailing and the quality of self-collected specimens was excellent. A positive test result was found in 22.1% overall and in 15.7% of self-reported non-vaccinated subjects. Among Canadian-born, non-vaccinated individuals, the positive test rate increased progressively from 1.2% (95% CI: 0-6.3) in 20-24 year olds to 16.4% (95% CI: 9.5-23.3) in 35-39 year olds. Antibody prevalence was higher among Canadian-born non-immunized 20-29 year olds who reported travel to developing countries (33.3%, 95% CI: 11.6-55.1) than in non-travellers (2.5%, 95% CI: 0.7-6.2). Conclusions: Mail-based population surveys of infection markers in oral fluid are feasible provided an appropriate sampling frame is used. This survey revealed a high anti-HAV antibody prevalence in young Canadian adults, increasing with age and travel to developing countries.