It has been suggested that moderate consumption of ethanol and wine has a protective effect on human health. Animal models used to date for alcohol consumption can not mimic real situations in humans because the consumption is forced and/or excessive. The present study proposes to determine the effects of a voluntary and ad lib consumption model more similar to that of human behavior. Male Wistar rats had free access to either standard diet and water or the same diet plus red wine, sweet wine, or a solution equivalent to red wine (13.5% ethanol) or to sweet wine (20% ethanol + 130 g/L sucrose) for 30 days or 6 months. Daily wine consumption was 15.8 +/- 0.9 and 2.0 +/- 0.2 ml/day for sweet and red wines, respectively. The consumption of each of the alcoholic solutions was similar to that of the wine they were simulating. Drinking wine or ethanol did not affect food and water intakes or growth rate. Plasma metabolites were not substantially affected by consumption of wine or ethanol. Although moderate and high wine consumption did not change the activity of plasma marker enzymes of tissue damage, the consumption of the 2 alcoholic solutions caused a long-term increase in the activity of aspartate aminotransferase. It seems that wine consumption protects the organism from hepatic lesions induced by ethanol alone.