The effects of the Lps gene on the development of experimental ulcerative colitis were studied in two genetically different mouse strains: C57B1 and C3H. Acute colitis was induced by adding 3% dextran sulfate sodium (DSS) to the drinking water for a 7-day (C57B1 and C3H) or a 10-day (C57B1) experimental period. Although the DSS treatment initiated the same type of morphological changes in the colon in all groups of mice, an earlier onset and persistent intestinal bleeding occurred in the Lpsn mice (sensitive to lipopolysaccharide, LPS) in comparison with the Lpsd mice (hyporesponsive to LPS). Rectal bleeding appeared on day 7 in 90% of the Lpsn compared to 13% of the Lpsd mice (p < 0.0001). In C57B1 mice, followed for three additional days, 50% of the Lpsn mice died and the surviving animals showed as well as rectal bleeding a large number of Gram-negative bacteria in the liver and spleen. In contrast, the Lpsd mice of the C57B1 strain appeared unaffected by the treatment, although a transient rectal bleeding occurred in 90% on day 8. Also, significantly fewer Gram-negative bacteria were found in the liver and spleen. Even though significantly increased serum endotoxin levels were seen in all DSS-treated groups compared to controls on day 7, the serum levels of TNF alpha were significantly increased only in the Lpsn mice. In DSS-induced colitis the Lpsn genotype conferred on the mice an increased LPS susceptibility, resulting in an augmentation of the inflammatory response to Gram-negative bacteria and their endotoxins. The results suggest that LPS-induced host effector mechanisms significantly enhanced the intestinal bleeding, systemic inflammatory response, and mortality in mice with DSS-induced colitis. In addition, the host defense against the invading and systemically spread bacteria most probably involved additional genes.