The waterborne transmission of hepatitis A virus (HAV), the main cause of acute hepatitis, is well documented. Recently, two ISO proposals for sensitive determination of this pathogen by RTqPCR in water and food have been published (ISO/TS 15216-1 and ISO/TS 15216-2), and could enable the formulation of regulatory standards for viruses in the near future. However, since detected viral genomes do not always correlate with virus infectivity, molecular approaches need to be optimized to better predict infectivity of contaminated samples. Two methods involving the use of propidium monoazide (PMA), with or without Triton X-100, prior to RTqPCR amplification were optimized and adapted to infer the performance of infectious viral inactivation upon two different water treatments: free chlorine and high temperature. Significant correlations between the decrease of genome copies and infectivity were found for both inactivation procedures. The best procedure to infer chlorine inactivation was the PMA-RTqPCR assay, in which 1, 2 or 3-log genome copies reductions corresponded to reductions of infectious viruses of 2.61 ± 0.55, 3.76 ± 0.53 and 4.92 ± 0.76 logs, respectively. For heat-inactivated viruses, the best method was the PMA/Triton-RTqPCR assay, with a 1, 2 or 3-log genome reduction corresponding to reductions of infectious viruses of 2.15 ± 1.31, 2.99 ± 0.79 and 3.83 ± 0.70 logs, respectively. Finally, the level of damaged virions was evaluated in distinct types of water naturally contaminated with HAV. While most HAV genomes quantified in sewage corresponded to undamaged capsids, the analysis of a river water sample indicated that more than 98% of viruses were not infectious. Although the PMA/Triton-RTqPCR assay may still overestimate infectivity, it is more reliable than the RTqPCR alone and it seems to be a rapid and cost-effective method that can be applied on different types of water, and that it undeniably provides a more accurate measure of the health risk associated to contaminated waters.