We assessed the possible association of gastrointestinal cancers with cruciferous vegetables and mushrooms in a multicenter, hospital-based case-control study in an agricultural area of Japan. One hundred forty-nine cases and 287 controls for stomach cancer and 115 cases and 230 controls for colorectal cancer were matched by age, sex, and residential area. In stomach cancer, the protective effect of vegetables (consumption of total vegetable) was obscure, but it became clearer when we examined specific kinds of vegetables. Marginal associations were observed in the group of the highest consumption of Chinese cabbage (odds ratio [OR] = 0.61; 95% confidence interval [CI] = 0.35-1.07), broccoli (OR = 0.60; 95% CI = 0.34-1.08), Hypsizigus marmoreus (Bunashimeji) (OR = 0.57; 95% CI = 0.31-1.04) and Pholita nameko (Nameko) (OR = 0.56; 95% CI = 0.30-1.06). In colorectal cancer, we observed decreased risks from the highest tertile of total vegetables (OR = 0.22; 95% CI = 0.08-0.66) and low-carotene-containing vegetables (OR = 0.28; 95% CI = 0.08-0.77), and inverse associations were observed in the group of the highest consumption of broccoli (OR = 0.18; 95% CI = 0.06-0.58). Although the sample size was limited, subgroup analyses showed that the associations differed with the histopathological subtype. These findings suggest that cruciferous vegetables decrease the risk of both stomach and colorectal cancer, and that mushrooms are associated with a decreased risk of stomach cancer.