It has been difficult to evaluate the protective efficacy of vaccine candidates against shigellosis, a major form of bacillary dysentery caused by Shigella spp. infection, because of the lack of suitable animal models. To develop a proper animal model representing human bacillary dysentery, guinea pigs were challenged with virulent Shigella flexneri serotype 2a (strains 2457T or YSH6000) or S. flexneri 5a (strain M90T) by the intrarectal (i.r.) route. Interestingly, all guinea pigs administered these Shigella strains developed severe and acute rectocolitis. They lost approximately 20% of their body weight and developed tenesmus by 24 h after Shigella infection. Shigella invasion and colonization of the distal colon were seen at 24 h but disappeared by 48 h following i.r. infection. Histopathological approaches demonstrated significant damage and destruction of mucosal and submucosal layers, thickened intestinal wall, edema, erosion, infiltration of neutrophils, and depletion of goblet cells in the distal colon. Furthermore, robust expression of IL-8, IL-1beta, and inducible NO synthase mRNA was detected in the colon from 6 to 24 h following Shigella infection. Most importantly, in our new shigellosis model, guinea pigs vaccinated with an attenuated S. flexneri 2a SC602 strain possessing high levels of mucosal IgA Abs showed milder symptoms of bacillary dysentery than did animals receiving PBS alone after Shigella infection. In the guinea pig, administration of Shigella by i.r. route induces acute inflammation, making this animal model useful for assessing the protective efficacy of Shigella vaccine candidates.