An in vivo model of ethanol ingestion in rats was used to examine tumor necrosis factor-alpha production after intravenous injection with lipopolysaccharide or saline solution. Four groups of 125-gm male Sprague-Dawley rats were given one of the following four diets: liquid ethanol diet (ethanol, 36% of calories), liquid control diet, chow ad libitum or control liquid diet pair-fed to match calories consumed by ethanol-fed rats. After 6 wk of diet, all rats were injected with 1 mg/kg lipopolysaccharide or 0.9% saline. AST concentrations in the ethanol-lipopolysaccharide group (388 +/- 54 U/ml) were significantly increased compared with those in control-saline, ethanol-saline and control-lipopolysaccharide groups (166 +/- 23, 166 +/- 18, 219 +/- 47; p < 0.01). Serum tumor necrosis factor-alpha concentrations for the ethanol-LPS group (3,990 +/- 624 pg/ml) were increased compared with those in control-saline (87 +/- 18), ethanol-saline (68 +/- 24) and control-LPS (695 +/- 165) groups (p < 0.001). A strong correlation was seen between serum tumor necrosis factor-alpha and AST concentrations (r = 0.91, p < 0.001). Treatment with lipopolysaccharide also increased transcriptional levels of tumor necrosis factor-alpha-specific mRNA from hepatic Kupffer cells isolated from rats fed the long-term ethanol diet by a factor of 3 compared with control rats. From these data, we conclude that long-term ethanol administration sensitized hepatic Kupffer cells to secrete high levels of tumor necrosis factor-alpha after lipopolysaccharide injection. Increased serum tumor necrosis factor-alpha concentrations correlated directly with increased levels of serum transaminase, which may have reflected hepatic injury.