Abstract Background Sapovirus (SaV) is a pathogen of acute gastroenteritis, and contains a positive-sense single-stranded 7.5 kb RNA genome. Characterization of SaV has been mainly performed with short nucleotide sequences, and the viral load has not been widely determined. Objectives To characterize SaV strains from two outbreaks in Nagano Prefecture, Japan, using long nucleotide sequences and to measure the viral load in feces. Study design Approximately 2.3 kb of the 3′ terminus of the genome corresponding to the entire capsid gene, ORF2 gene, and 3′ untranslated region were amplified with semi-nested RT-PCR followed by sequencing analysis. The copy numbers of the SaV genome were determined with real-time RT-PCR. Results In Outbreak 1, SaV strains belonging to genogroup I (GI) were detected from seven symptomatic nursery children, in which six SaV isolates had identical nucleotide sequences while one had a single synonymous nucleotide substitution. In Outbreak 2, two similar GIV SaV sequences were detected, in which three nucleotide differences accompanying two non-synonymous substitutions were observed between symptomatic high school students and asymptomatic food handlers at a hotel. The cDNA copies were 1.36 × 10 7 to 1.05 × 10 11 ( n = 7), and 5.05 × 10 6 to 1.27 × 10 10 ( n = 6) per gram of stool specimens in the two outbreaks. Conclusions The nucleotide sequence covering the 3′ terminal 2.3 kb of the genome is useful for better characterization of the SaV strains. In addition, we found for the first time adults who secreted SaV with high viral loads without gastroenteritis symptoms.