There is still insufficient knowledge of the distribution of drinking habits in human populations, and more descriptive surveys are needed. Both prospective and retrospective epidemiological studies indicate that alcohol consumption is a cancer hazard. Prospective studies on excessive drinkers have shown an increased risk for cancer of the mouth, pharynx, larynx, esophagus, liver, and lung. Retrospective studies have confirmed this excess risk. For cancers of the buccal cavity, pharynx, larynx, and esophagus, the effect of drinking has been shown to be associated with the effect of smoking. In the case of esophageal cancer, these two effects are independent, and the observations made are consistent with a multiplicative model. Primary liver cancer is also associated with alcohol consumption, probably by a less direct action; the importance of the impact of alcohol on primary liver cancer is probably underestimated. Animal experiments have not shown that ethanol alone has a carcinogenic effect, and the mechanisms by which alcoholic beverages act on humans remain unknown. The proportion of cancer cases at sites known to be associated with alcohol consumption is approximately 8% in most population groups in the United States. This indicates that a sizeable proportion of cancers is potentially preventable if appropriate action is taken.