The age-corrected mortalities from cancer at 17 sites were correlated with the consumptions of 12 major food items and the apparent consumptions of alcoholic beverages, cigarettes, beer, coffee, tea and solid fuel, in different countries. Significant direct and inverse correlations with food consumption variables were observed: the results are in essential accord with those of other workers employing different methods of statistical analysis. A modification of the diet in western industrialized nations is proposed as a means of lowering the human cancer risk. For cancers of the mouth and neck, and for liver in males, significant associations with the wine alcohol consumption were observed. Statistical evidence for liver cirrhosis as liver-cancer, and for stomach ulcer as stomach-cancer-predisposing conditions was obtained.