The objective is to study the possible mechanism by which lactic acid bacteria ( LAB) prevent alcoholinduced steatohepatitis in rats. A total of 25 Wistar rats were divided into three groups: a LAB- fed group, an alcoholtreated group and a control group. Both the LAB- fed group and the alcohol- treated group received alcohol ( 10 g kg- 1 per day) orally for up to 5 days ( 125 h). Before exposure to alcohol, the LAB- fed group were first treated daily with 1.5 ml/ 100 g of a mixture comprising 4 9 10 10 ml- 1 of Lactobacillus acidophilus and 2.5 9 10 7 ml- 1 of Bifidobacterium longum, while the control group was treated with normal saline only. Biochemical data, alcohol dehydrogenase ( ADH) activity and histology of the liver and stomach were evaluated. The ADH activity in the LAB mixture was 3.52 +/- 0.45 lmol mg- 1 protein ( 10 9 CFU ml- 1), and was dose- dependent. By 30 min after taking alcohol, serum alcohol concentrations were 514.24 +/- 80.21 mu g ml- 1 in the LAB- fed group and 795.15 +/- 203.45 mu g ml- 1 in the alcohol- treated group ( P < 0.005). Serum alcohol concentrations were reduced by 48% ( P < 0.01) in the LAB- fed group, but by only 4% in the alcohol- treated group ( P[ 0.05) 120 min after oral intake of alcohol. The blood levels of endotoxin, AST and ALT were improved in the LAB- fed group compared to the alcohol- fed group ( P < 0.01). All alcohol- treated rats showed moderate to severe steatohepatitis, but the LAB- fed rats showed almost normal histology or very slight lesions only. In conclusion, LAB decreased the alcohol concentration in the blood by increasing the first- pass metabolism in both the stomach and the liver, and effectively protected against alcohol- induced gastric and liver injury. It is interesting to note that the protection was more effective in the liver.