Changes in the estimated proportion of liver cirrhosis deaths attributable to alcoholism in the United States from 1940 to 1980 are reviewed. The value of this proportion from 1940 in the original Jellinek Alcoholism Estimation Formula, through Formula modifications, to 1980 use are presented; the rationale for various changes in its value are discussed. Essex County, New Jersey, USA decedents in 1984, aged 25-years or older, whose underlying cause of death was specified alcoholic cirrhosis and those who died of cirrhosis without mention of alcohol are analyzed for differences in background and post-mortem characteristics. Some appreciable proportion of cirrhosis deaths without mention of alcohol is considered to be attributable to alcoholism. Background and post-mortem differences between persons whose underlying cause of death is certified as cirrhosis with and without mention of alcohol suggest some bases for the under-reporting of specified alcoholic cirrhosis mortality.