Due to the ease of collection, oral fluid is being investigated as an alternative to serum for diagnostic and epidemiological purposes. However, for prospective studies involving hepatitis A virus (HAV) RNA detection, a standard methodology must be developed. In the present study, nested RT-PCR and real-time PCR were optimized and evaluated for HAV detection and quantification, using oral fluid from healthy volunteers (n=20) and paired serum/oral fluid samples from individuals involved in a hepatitis A outbreak (n=78). Using nested RT-PCR, HAV RNA was detected in 50% of oral fluid and in 42% of serum samples from acute cases, as well as in 12% of all samples from cases without IgM and total anti-HAV. Using real-time PCR, HAV RNA was detected in 61% of oral fluid and in 71% of serum samples from acute cases, as well as in 17 and 12%, respectively, from patients without HAV markers. Mean viral loads were 1.7+/-3.24 x 10(3)copies/ml in oral fluid and 2.8+/-6.46 x 10(3)copies/ml in serum. Although nested RT-PCR and real-time PCR both detected HAV RNA in oral fluid, real-time PCR was more sensitive. Oral fluid sample testing could be used as a noninvasive method of detecting HAV RNA during HAV outbreaks.