The relation between selected micronutrients and oral and pharyngeal cancer risk was investigated using data from a case-control study conducted between January 1992 and November 1997 in Italy and Switzerland. Cases were 754 incident, histologically confirmed oral cancers (344 of the oral cavity and 410 of the pharynx) admitted to the major teaching and general hospitals in the study areas. Controls were 1,775 subjects with no history of cancer admitted to hospitals in the same catchment areas for acute, non-neoplastic diseases. Dietary habits were investigated using a validated food-frequency questionnaire. Odds ratios (ORs) were computed after allowance for age, sex, center, education, occupation, body mass index, smoking and drinking habits and non-alcohol energy intake. Micronutrients were analyzed both as continuous variables and in quintiles. In the former case, the unit was set to 1 SD of the distribution of controls. ORs for the continuous analysis were 0.95 for retinol, 0.61 for carotene, 0.91 for lycopene, 0.83 for vitamin D, 0.74 for vitamin E, 0.63 for vitamin C, 0.82 for thiamine, 0.87 for riboflavin, 0.59 for vitamin B6, 0.61 for folic acid, 0.62 for niacin, 0.91 for calcium, 0.88 for phosphorus, 0.65 for potassium, 0.82 for iron, 0.67 for non-alcohol iron and 0.89 for zinc; the 95% confidence interval excluded one for carotene, vitamin C and E, thiamine, vitamin B6, folic acid, niacin, potassium and iron. ORs were similar for the 2 sexes and in strata of age. When the combined intake of vitamins C and E and carotene was considered, the protective effect of each nutrient was more marked or restricted to subjects with low intake of the other 2. The association with vitamin C and carotene was independent of smoking and drinking habits, while that with vitamin E was less evident in those heavily exposed to alcohol or tobacco. In general, the more a micronutrient was correlated to total vegetable and fruit intake, the stronger was its protective effect against oral cancer.