The effects of long-term moderate or "social" alcohol consumption (10-80 g daily intake) on the incidence of features of alcoholic liver disease (ALD) were delineated in a consecutive autopsy series of 210 males. The subjects' daily intake, as well as duration of alcohol consumption, was determined by an interview with the spouse or a close acquaintance and compared with semiquantitative histological scores for stage of ALD. No significant increase in the incidence of features of ALD could be related to all-year daily intake of ethanol below 40 g (40 g equals 1.1 liter of beer, 0.44 liter of wine, and 0.11 liter of spirits). However, daily intake between 40-80 g increased relative liver weight on average 3.1 g/kg of body weight (p < 0.02), the frequency of fatty liver from 11.7 to 47.2% [relative risk (RR) = 4.4], and the frequency of mainly slight alcoholic hepatitis up to 16.7% (RR = 7.5). The incidence of both bridging fibrosis and liver cirrhosis increased significantly (RR = 8.8) only when daily intake exceeded 80 g. Amounts of ethanol exceeding 80 g did not relate to further increases in incidence of bridging fibrosis or liver cirrhosis. These findings suggest that, in males, daily ingestion of ethanol below 40 g for a period of 25 years does not increase the risk of alcohol-related liver disease.(ABSTRACT TRUNCATED AT 250 WORDS)