Despite the increasing numbers of studies investigating hepatitis A diagnostic through saliva, the frequency and the pattern of hepatitis A virus (HAV) markers in this fluid still remains unknown. To address this issue, we carried on a longitudinal study to examine the kinetics of HAV markers in saliva, in comparison with serum samples. The present study followed-up ten patients with acute hepatitis A infection during 180 days post diagnosis (dpd). Total anti-HAV was detected in paired serum and saliva samples until the end of the follow-up, showing a peak titer at 90th. However, total anti-HAV level was higher in serum than in saliva samples. This HAV marker showed a probability of 100% to be detected in both serum and saliva during 180 dpd. The IgM anti-HAV could be detected in saliva up to 150 dpd, showing the highest frequency at 30th, when it was detected in all individuals. During the first month of HAV infection, this acute HAV marker showed a detection probability of 100% in paired samples. The detection of IgM anti-HAV in saliva was not dependent on its level in serum, HAV-RNA detection and/or viral load, since no association was found between IgM anti-HAV positivity in saliva and any of these parameter (p>0.05). Most of the patients (80%) were found to contain HAV-RNA in saliva, mainly at early acute phase (30th day). However, it was possible to demonstrate the HAV RNA presence in paired samples for more than 90 days, even after seroconversion. No significant relationship was observed between salivary HAV-RNA positivity and serum viral load, demonstrating that serum viral load is not predictive of HAV-RNA detection in saliva. Similar viral load was seen in paired samples (on average 104 copies/mL). These data demonstrate that the best diagnostic coverage can be achieved by salivary anti-HAV antibodies and HAV-RNA tests during 30-90 dpd. The long detection and high probability of specific-HAV antibodies positivity in saliva samples make the assessment of salivary antibodies a useful tool for diagnosis and epidemiological studies. The high frequency of HAV-RNA in saliva and the probability of detection of about 50%, during the first 30 dpd, demonstrate that saliva is also useful for molecular investigation of hepatitis A cases, mainly during the early course of infection. Therefore, the collection of saliva may provide a simple, cheap and non-invasive means of diagnosis, epidemiological surveys and monitoring of hepatitis A infection purposes.