Classification of individuals by their vitamin C intake was investigated in 493 control subjects from a cervical cancer case-control study. The influence of dietary and supplemental sources of vitamin C, as well as demographic and life-style factors, on serum ascorbic acid were examined. Usual dietary intakes of vitamin C were determined from a food frequency questionnaire and recent intakes from a 24-hour recall taken at the time of blood collection. Vitamin supplement information was obtained at both times. In a regression analysis, the factors found to predict serum ascorbic acid were total recent vitamin C intake, an indicator variable for supplement use, body mass index, number of cigarettes smoked per day, race, education, and age. Higher levels of serum ascorbic acid were found among older nonsmoking highly educated leaner white women. Consideration of supplements, in addition to dietary sources of vitamin C, improved correlation coefficients between serum ascorbic acid and usual vitamin C intake from 0.19 to 0.32 and between serum ascorbic acid and recent intake from 0.36 to 0.56. Furthermore, whereas only a twofold difference between the first and fourth quartiles of serum ascorbic acid was observed using recent dietary vitamin C without supplements, this range increased to sixfold with addition of supplement data. Epidemiological studies should consider use of total vitamin C intakes from supplement and food sources to permit accurate classification of individuals.