We reviewed data from six cohort studies and approximately 40 case-control studies on the relation between selected aspects of diet and the risk of oral and pharyngeal cancer. Fruit and vegetables were inversely related to the risk: the pooled relative risk (RR) for high vegetable consumption was 0.65 from three cohort studies on upper aerodigestive tract cancers and 0.52 from 18 case-control studies of oral and pharyngeal cancer; corresponding RRs for high fruit consumption were 0.78 and 0.55. beta-carotene, vitamin C and selected flavonoids have been inversely related to the risk, but it is difficult to disentangle their potential effect from that of fruit and vegetables. Whole grain, but not refined grain, intake was also favorably related to oral cancer risk. The results were not consistent with reference to other foods beverages, and nutrients, but it is now possible to exclude a strong relation between these foods and oral and pharyngeal cancer risk. In western countries, selected aspects of diet may account for 20-25% of oral and pharyngeal cancer, and the population attributable risk increases to 85-95% when tobacco and alcohol consumption are also considered.