Two distinct hepatitis E virus (HEV) isolates, designated HE-JI3 and HE-JI4, were identified in a single patient with acute hepatitis in Japan, who had not travelled abroad. The HEV load of HE-JI3 at admission was 10(2) copies/ml, but that of HE-JI4 was tenfold higher at 10(3) copies/ml. The viraemia of HE-JI4 persisted for up to 16 days from admission, whereas HE-JI3 disappeared at 9 days after admission. The entire nucleotide sequence of the HE-JI4 isolate and partial nucleotide sequences of open reading frames (ORFs) 1 and 2 of the HE-JI3 isolate were determined. The full-length nucleotide sequence of HE-JI4 consisted of 7171 nucleotides excluding the poly(A) tail and contained ORF1 encoding 1684 amino acids, ORF2 encoding 671 amino acids and ORF3 encoding 114 amino acids. Sequence and phylogenetic analyses of the HEV genomes indicated that HE-JI4 was most closely related to an HEV isolate (T1) of genotype IV with the same strategy for translation of ORF2 and ORF3, but which differed from it by 16.5% over the entire genome. The HE-JI3 isolate showed the highest nucleotide identity (88.6-95.1%) to the genotype III HEVs, having higher identity to human and swine HEV isolates from the United States (US1, US2 and swUS1) than to those reported thus far from Japan (JRA1 and swJ570). The two co-infecting strains of HE-JI3 and HE-JI4 identified from the single patient shared only 80.1% nucleotide identity. These results indicate that multiple genotypes of HEV co-circulate in Japan, and that genotype IV comprises a remarkably heterogeneous group of HEVs.