Per capita consumption of alcohol rose steadily in the U.K. from 1970 to 1979, but fell by 11% between 1979 and 1982. This fall in consumption was followed by a 19% fall in first admissions for alcohol dependence, a 16% fall in drunkenness convictions, a 7% fall in drinking and driving convictions and a 4% fall in cirrhosis mortality. Between 1970 and 1982 there were highly significant (P less than 0.01) correlations between per capita consumption and convictions for drunkenness and drinking and driving, first admissions to hospital for alcohol dependence, and mortality from cirrhosis, pancreatitis and cancer of the oesophagus. These findings add weight to the argument that per capita consumption is the crucial variable determining the magnitude of the burden imposed on the community by the harmful effects of excessive drinking.