Comparative epidemiological studies with ecological and case-control approaches in high- and low-epidemic areas of China have provided us with much evidence with regard to risk and benefit in the environment. To clarify how dietary factors are involved in esophageal and stomach cancer development, we performed a case-control study in a low-epidemic area, and compared the findings with those obtained earlier for a high-epidemic area for stomach cancer in the same Jiangsu Province, China. We recruited 199 and 187 cases with esophageal and stomach cancers, respectively, and 333 population-based common controls. Odds ratios (ORs) for esophageal and stomach cancers were calculated with adjustment for potential confounding factors, using an unconditional logistic model. Current and former smoking elevated the OR for esophageal cancer, along with high intake of pickled vegetables and broiled meat, while decreased ORs were observed for frequently consumed raw vegetables and garlic. With regard to stomach cancer, ORs were increased with frequent consumption of salty fish, leftover gruel, and broiled meat, and lowered by snap bean consumption. The present risk factors were common to the previously obtained results in the high-epidemic area, and similarly distributed in each general population. While more protective factors were observed in the high-epidemic area, their penetrance was much greater in the low-epidemic area. The present study thus suggests that frequent vegetable and garlic consumption contributes to low mortality rates for esophageal and stomach cancers in a low-epidemic area, counteracting similar exposure levels for risk factors as in the high-epidemic area.