Bovines are recognized as an important reservoir of Shiga toxin-producing Escherichia coli (STEC). Although STEC strains are significant foodborne pathogens, not all of the STEC held by cattle are pathogenic, and which type of STEC that will become epidemic in humans is unpredictable. Information about the prevalence of serotype and virulence gene distribution in beef cattle is insufficient to develop monitoring and controlling activities for a food safety and security program. Thus, this study investigated the prevalence of O157 and non-O157 STEC in Japanese beef cattle and characterized the isolates by the type of O antigen and several virulence markers to help predict the pathogenicity. In this study, 64.2% (176 of 274) of enrichment cultures of fecal samples collected from an abattoir and farms were stx1 and/or stx2 positive by PCR. STEC strains were isolated from 22.1% (39 of 176) of the positive fecal samples, and these isolates represented 17 types of O antigen (O1, O2 or O50, O5, O8, O55, O84, O91, O109, O113, O136, O150, O156, O157, O163, O168, O174, and O177). Two selective media targeting major STEC groups, cefixime-tellurite sorbitol MacConkey agar and CHROMagar O26/O157, allowed isolation of a variety of STEC strains. The most frequently isolated STEC was O113 (8 of 39), which has previously been reported as a cause of foodborne infections. Although most of the O113 STEC isolated from infected patients possessed the enterohemolysin (hlyA) gene, none of the O113 STEC cattle isolates possessed the hlyA gene. The second most common isolate was O157 (6 of 39), and all these isolates contained common virulence factors, including eae, tir, lpf1, lpf2, and hlyA. This study shows the prevalence of O157 and non-O157 STEC in Japanese beef cattle and the relationship of O antigen and virulotypes of the isolates. This information may improve identification of the source of infection, developing surveillance programs or the current understanding of virulence factors of STEC infections.