Hepatobiliary inflammation and other extraintestinal manifestations accompany certain intestinal disorders, perhaps because of proliferation or enhanced transport of luminal bacteria or their phlogistic cell-wall components. Using jejunal self-filling blind loops to create small bowel bacterial overgrowth, we compared biochemical and histological evidence of hepatic inflammation in 3 rat strains chosen for their variable inflammatory responses to bacterial cell wall polymers. Lewis and Wistar rats developed weight loss, hepatomegaly, and hepatic inflammation 4 and 12 wk, respectively, after creation of SFBL. Plasma aspartate aminotransferase levels in Lewis rats 4 wk (578 +/- 77 U/L) and Wistar rats 12 wk (220 +/- 35 U/L) after self-filling blind loops were significantly greater than in rats with self-emptying blind loops (112 +/- 24 U/L, p less than 0.001; 104 +/- 22, p less than 0.05) or sham-operated Lewis (84 +/- 24, p less than 0.001) or Wistar (78 +/- 10, p less than 0.001) rats. Randomized comparison using a histology grading score showed abnormalities that paralleled aminotransferase values. Lewis and Wistar rats with self-filling blind loops had hepatic injury with bile duct proliferation, fibrosis and acute and chronic periportal and focal parenchymal inflammation. Lewis and Wistar rats with self-emptying blind loops developed occasional mild histologic lesions. 50% of Lewis rats with self-filling blind loops for 4 wk died compared with only 15% in other groups. However, Buffalo rats with self-filling blind loops developed no weight loss, hepatomegaly, or hepatic injury. Anaerobic cultures of blood, peritoneum and liver were negative in all strains. Diet-restricted, sham-operated Wistar rats with weights similar to the Wistar rats with self-filling blind loops did not develop histologic abnormalities or elevated aminotransferase levels (76 +/- 31 U/L). These results show that experimental small bowel bacterial overgrowth causes significant hepatic inflammation leading to fibrosis in susceptible rat strains. Caloric deprivation and hepatic bacterial invasion are not etiologically responsible. We suggest that bacterial cell wall polymers or other bacterial toxins from the blind loop cause hepatic lesions in genetically susceptible hosts.