To investigate confounding factors in the relation of passive smoking to diseases, we compared the dietary intake of passive smokers, non-smokers without passive smoke exposure, and smokers. The subjects were female respondents to a baseline survey, which was conducted as part of a collaborative cohort study in a rural area. Of the subjects, 101 females were smokers. A total of 1978 female non-smokers answered the question about passive smoking exposure at home, including 1,392 (70.4%) passive smokers and 586 (29.6%) non-passive smokers. Among these three groups, the dietary intake of 36 foods (frequency and amount) was compared by odds ratios calculated with a logistic regression model. The percentages of subjects reporting frequent intake of milk or milk products, carrot or pumpkin, tomatoes, oranges, and fruits except oranges, were significantly lower in passive smokers than in non-passive smokers (OR = 0.80, 0.74, 0.80, 0.77, 0.79). On the other hand, more subjects in passive smokers reported frequent or large intake of pork, salt pickled vegetables, soy sauce pickled foods, soft drinks, coffee, and moso soup (OR = 1.38, 1.53, 1.32, 1.73, 1.30, 1.33). The dietary pattern of passive smokers was similar to that of smokers. In conclusion, in this study, passive smokers had different dietary patterns from non-passive smokers'. In future research regarding disease with passive smoking exposure, dietary factors should be considered as a confounder.